1

THE GOLDEN AGE OF THE SWEDISH MODEL

The Coherence Between Capital Accumulation and

Economic Policy in Sweden in the Early Postwar Period

Lennart ErixonDepartment of EconomicsStockholm UniversityOctober 1996

2

Content

Introduction

Chapter 1 - Altenative Views of the Swedish Golden Age

Chapter 2 - The Swedish Model

Chapter 3 - The Swedish Growth Engine

3.1 Raw Material Industry - A Nordic Component of the Swedish Model

3.2 Outward Large Companies in Enginering - An Un-Nordic Component ofthe Swedish Model

3.2.1 The First Innovation Wave3.2.2 The Second Innovation Wave3.2.3 The Early Foreign Orientation of Swedish Companies

Chapter 4 - In What Way Are the Leading Industries Leading?

4.1 Shares of Production and Employment in Swedish Manufacturing

4.2 Net-Works and Multipliers

4.3 Wage Leadership

4.4 The Production and Diffusion of Knowledge by Large Engineering Companies

4.4.1 The Fundamental Innovation System4.4.2 The Regulated Innovation System

Chapter 5 - The Performance of the Swedish Model in the Golden Age

5.1 Threats, Opportunities and Reactions I - The Large Engineering Companies

5.2 Threats, Opportunities and Reactions II - The Raw Material Industries

Chapter 6 - The Coherence of the Swedish Model

6.1 The Embryo of the Swedish Model

6.2 Economic Policies in the Golden Age6.2.1 The Early Golden Age6.2.2 The Late Golden Age6.2.3 “What´s Good for Volvo is Good Also for Sweden”

Chapter 7 - Black Clouds at the Horizon - Tensions in the Swedish Model

3

7.1 Internal Conflicts in the Model7.2 An Exhaustion of Growth Potentials of the Swedish Model7.3 The Social Price of Outwardness and Specialization7.4 Locking-In Effects

7.4.1 The Early Golden Age7.4.2 The Late Golden Age

The Swedish Growth Model - A Summary

Literature

Tables

Introduction

4

The Swedish model seems to be in trouble. Sweden was haunted in the early 1990s bydeep and frequent financial and foreign exchange crisis. GDP growth was negative forthree years 1991-1993, which is unique for the postwar period not only for Swedenbut also for the OECD countries if Finland is excluded. Simultaneously, Sweden´spublic deficit (per cent of GDP) grew larger than in other OECD countries with theexception of Greece. To raise confidence for Sweden in international financial marketsand adjust to the European Monetary Union, politicians have agreed to propose drasticreductions in public transfers.

While previously famous for its commitment to full employment, Sweden experienceda sharp rise in the rate of unemployment almost to the EU average level from 1990 to1993. Open unemployment was reduced only marginally in the mid 1990s in spite of aSocial Democratic government from Autumn 1994 and a recovery in GDP growth.The current rate of unemployment (in October 1996) is almost 8 per cent.

The Swedish case has fascinated social scientists and politicians abroad. Sympatheticobservers have considered the Swedish model as a combination of the best elementsfrom both socialism and capitalism, combining high economic growth with fullemployment and equality-orientated policies. From this vantage point, the currentmacroeconomic problems are either transitional or they reflect a betrayal of theprinciples of the original model.

Sceptics, on the other hand, have seen Sweden as the ultimate example ofEurosclerosis, leading to a crisis of the Social Democratic welfare state. In such aview, the present macroeconomic difficulties are unavoidable consequences of thepriorities and arrangements embedded in the model.

A second sceptical group can be distinguished which is not necessarily hostile to “theSwedish model” as such, but rather questions its validity. The group argues, byreference to the Swedish crisis in the 1990s, that it is impossible to maintain fullemployment, income equality and a large public sector in a world of transnationalfirms, borderless capital markets, EU integration, strong international competition anddeflationary economic policies in major OECD countries.

Sweden had experienced a “light” and a “grey” period before the darkness of the 1990saccording to the prevailing objectives of economic policy. The light period coincideswith the Golden Age in Angus Maddison´s Phases of Capitalist Development [1982],thus it starts when World War II is over and ends with the first oil crisis. Productivitygrowth was high then in Sweden, not only historically (as production growth) but alsoin an international perspective if countries which were seriously hit by the war areignored. The rate of unemployment was low and the potential trade-off between lowunemployment and inflation satisfactory solved, especially in the early 1970s.Fluctuations in GDP were milder in Sweden than in most other OECD countries (seetable 1). The Swedish success is completed by the fact that wage and incomedifferentials were smaller than in any other OECD country at the end of the GoldenAge. Thus, Sweden confirmed during the Golden Age that growth can be combinedwith equality.

5

The grey period in Sweden starts at the first oil crisis and ends with the downturn inthe early 1990s. In this period, production and productivity growth in Sweden becamemuch lower than earlier. A similar decline occurred in other OECD countries. But itseems that Swedish growth was lower than in both comparable and competingcountries (Erixon [1991]). There were also strong fluctuations in the Swedish GDP.Furthermore, inflation became higher in Sweden than earlier and also higher than in theEU and OECD (on average) during “the long peak” in the 1980s and the early 1990s.Inflation had unintended effects on income distribution both before and after taxes inSweden. In addition, wage differentials, inter alia between the sexes, began to widenin the 1980s.

But the macroeconomic outcome for Sweden after the first oil crisis was notunequivocally depressing. The country managed to keep “full employment” when theEU countries, including Denmark, entered the road to permanent mass unemployment.In addition, Sweden obtained a better balance between unemployment and inflationthan most other OECD countries until the late 1980s. Finally, the fluctuations inSwedish GDP were not exceptional in the period under discussion.

This essay on Sweden will exclusively consider the light period, a period that is heresynonymous with the Swedish Golden Age.1 Sweden is evaluated in terms of hergrowth performance. My emphasize is laid on the Swedish growth engine, the rawmaterial industries and the large transnational companies in engineering. I will uncoverthe causes behind and macroeconomic effects of their success in the Golden Ageparticularly in relation to Social Democratic policies. This relation is adressed onlyperipherally or simply ignored by most of the literature on “the Swedish model”.2

My motivation for a paper on Sweden in the Golden Age only is simple - the recentfailures of “the Swedish model” are better understood in the light of its previoussuccess. I will stress not only the achievements of the Swedish growth model but alsoits limits and negative effects. The Swedish success in 1950s and the 1960s was eithertemporary or paved the way to the following failures.

I will use a comparative approach to separate institutions and developments which aretypically Swedish. Sweden is continuously compared to other OECD countries,primarily the other Nordic countries. My essay is a by-product of a work on economicpolicies in the Nordic countries in the 1980s and the 1990s (Mjøset et al [1997]).

Chapter 1 - Alternative Views of the Swedish Golden Age

1 In an international perspective, the notion of a Swedish Golden Age is more accurate for a periodwhich lasted from 1870 to World War I. In this period, growth in GDP per hour and per head inSweden was higher than in any other country covered by Maddison´s study. The Swedish growth rateswere also among the highest in the interwar period (Maddison [1982, pp. 44-45 and 96-98]). Incontrast to other countries, growth was high in Sweden during both the 1920s and the 1930s.

2 There are a few exceptions, see Erixon [1985] and Mjøset et al [1986].

6

Three theories of the Swedish success in the Golden Age can be distinguished in socialscience - a liberal-corporatist theory, a social democratic theory and an “external”theory about favourable international circumstances. The theories correspond roughlyto the division of sceptics and friends of “the Swedish model” above. To some degree,however, the theories can be integrated.

The liberal-corporatist theory assumes that growth is impeded by public sectorexpansion, high taxes and compressed wage differentials (cf. Olson [1990, pp. 1-7]).Such phenomena will emerge through the claims from strong interest groups,particularly labour. High growth in the European welfare countries in the 1950s andthe 1960s is simply explained by the fact, that the size and negative effects of thecorporate state were still moderate. In the Swedish case, public sector growth did notaccelerate until the 1970s. Liberal theorists do admit that growth might be stimulatedby some corporatist tokens. According to Mancur Olson, encompassing organizationssuch as the LO, the Swedish TUC, have prevented price and trade regulations inSweden. Liberal economists have added, that negative wage effects of strong tradeunions are “internalized” in countries like Sweden by centralized bargaining (Calmfors[1992, p. 14]).

The social democrats share the liberal-corporatist view that centralized wagenegotiations have promoted Swedish growth. On the other hand, social democraticresearchers deny that small income differentials, high taxes and public sector growthhave strong disincentive effects. They even claim that growth was stimulated in theGolden Age by “the Swedish model”. Solidaristic wage policy put ineffective firmsunder a cost pressure and labour market policies provided expanding firms with labour.Social democrats argue also, that a successful counter-cyclical fiscal and monetarypolicy in the Golden Age created stable conditions for private investments.3

According to the “external” view, growth conditions were extremely advantageous forSweden in the early postwar period. As specialized in raw materials and investmentgoods, Swedish industry was strongly favoured by the rebuilding of Europe.Moreover, its main competitors had not yet recovered from the war. Tradeliberalizations engendered new market opportunities for Swedish companies which alsohad great possibilities to catch up new American technologies. Finally,internationalization of financial and product markets was not far-reaching, thuspermitting an independent Swedish fiscal and monetary policy.

All theories of the Swedish Golden Age have weaknesses. The liberal-corporatisttheory ignores one side of Swedish corporatism - the strategic role of largetransnational companies, mainly in engineering. It hides, for instance, that theexpansion of leading industries was promoted by Social Democratic policies. A majorexplanation of the worser performance of Sweden in the 1970s and the 1980s is thatthe Swedish growth model became less effective and that political supports to leading

3 There are other possible elements of a social democratic growth theory. According to IngvarSvennilson, wage earners´ accept technical change to a larger extent and are more mobile at fullemployment (Svennilson [1954, pp. 24, 36 and 49]).

7

industries have stunted growth.4 Such a conclusion is also hidden by the socialdemocrats who either tone down the economic problems in Sweden in the grey (andblack) period or blame them on departures from “the Swedish model”. The fact thatthe large export companies are wage leaders in Sweden is disguised by both liberal-corporatist and social democratic theories where wage bargaining systems are put infocus.

In the “external” theory, Swedish growth is discussed in terms of good and bad luck orrelated to factors beyond the control of domestic actors. It can be objected thatadjustments of companies and politicians to external events ultimately determinegrowth. The external theory also obscures the fact that internationalization has been aprerequisite rather than a new condition for “the Swedish model” in the postwarperiod. Moreover, the possibilities for Sweden to pursuade independent monetarypolicies were limited already in the 1960s.

Hitherto, the meaning of the concept “the Swedish model” has been vague indeed. Itremains to give the concept a more precise account.

Chapter 2 - The Swedish Model 4 Mancur Olson´s assumption of minor trade and price regulations in Sweden is hard to reconcilewith a conclusion by the OECD secretariat that weak competition, primarily in food products,construction and services, is an important explanation of high inflation and the regress of relative percapita income in Sweden in the 1980s (OECD [1992a, ch. IV]).

8

There is no established definition of the Swedish model. It thus comes as no surprisethat views on its rise and decline are diverging. In principle, there are two ways ofdefining the Swedish model. The first one considers Sweden as a consensus model.References are often made to the postwar political unity as to stabilization policies andwelfare programmes. Furthermore, advocates of a consensus approach emphasize thestrong accord between capital and labour in Sweden originating in the late 1920s andthe 1930s. Here, the 1938 Saltsjöbaden agreement that labour conflicts should beresolved through negotiations and not by non-neutral legislation is an important event.The employers´ initiative to central wage agreements in the mid 1950s is anothermilestone in a history of Swedish consensus.

According to the consensus approach, the Swedish model was challenged by labour inthe mid 1970s. The proposals of wage earner funds questioned the prerogative ofprivate ownership. Furthermore, legislation on job security and participation brokewith the tradition of non-interventionist labour relations. Employers´ refusal from themid 1980s to participate first in central wage negotiations and then in corporatistpublic bodies confirms the collapse of the Swedish consensus model.

The definition of the Swedish model in terms of consensus is marred by someweaknesses. It is difficult to distinguish genuine Swedish features since similar socialaccords have been developed not only in other Nordic countries but also in continentalEuropean countries such as Germany and Austria. Besides, it is difficult tounequivocally interpret changes in economic policy in Sweden during the 1980s andthe 1990s as either an indication of or a break with the consensus model. Neworientations of economic policy have received an unanimous support from establishedpolitical parties but been rejected by trade unions.

I will prefer an alternative definition of the Swedish model where Sweden is seen as acountry of socio-economic balance between labour and the leading industrial sector(cf. Korpi [1983]). According to this view, the priorities and instruments/institutions ofa strong Social Democracy and (blue collar) trade union movement are consistent withthe growth of raw material industries and transnational firms. In fact, the Swedishmodel is a coherent system where sustainable growth and profitability of the leadingindustrial sector is a necessary condition for obtaining the goals of the labourmovement, and economic policy, in its turn, guarantees capital accumulation in thissector - see figure 1.

Figure 1: The Swedish Postwar Model

9

Actors/sectors Social Democratic and Export-OrientatedBlue-Collar Raw Material Indu-Union Strength stries and Large Trans-(Political, Ideological national Companies inand Theoretical) Engineering

Priorities Full Employment SustainableGeneral Welfare Growth and(Including Real ProfitabilityWage Growth)Equality

Means State Interventionismfor Full EmploymentGeneral Welfare PoliciesWork Life ReformsCentral Wage BargainingRedistributive Incomeand Wealth TaxesSupport to LeadingCompanies/Industries

The parliamentary, organizational and ideological strength of the Swedish reformistlabour movement in the postwar period has no correspondence in other democracies,not even in other Nordic countries.5 The Social Democrats led all governments until1976. The union density rate was, and still is, higher in Sweden than in other OECDcountries both among wage and salary earners (Visser [1989]). Today, the LOorganises almost all blue-collar workers in Sweden.

Labour strength resulted in a strong priority of full employment, redistribution andgeneral welfare in Sweden in the Golden Age.6 The LO had a strong influence oneconomic and welfare policies and on legislation to improve working conditions, jobsecurity and labour participation. Major working life reforms, however, were notdecided until the end of the Golden Age.

5 Norway is the country which comes closest to Sweden in these respects. But the union density ratehas decreased in Norway since the 1960s while it has inceased (even from a high level) in Sweden.Besides, there is no correspondence in Norway to the influence of the LO economists on Swedisheconomic policies in the Golden Age.

6 The generality of welfare policy is not a typical Swedish, or even Nordic, token. However, it ispossible, if Denmark is excluded, to define a Nordic welfare model which is general and wherecompensations in social insurance systems are related to income (Korpi & Palme [1993, pp. 139-140]). Sweden´s welfare reforms often became standards for other Nordic countries. For instance, theincremental pension fund system (the ATP system) was first introduced in Sweden, and then imitatedby Norway and Finland (although the pension funds in Finland were based on collective agreements).

10

Some misunderstandings about the Swedish model must be pointed out. Foreignobservers think that Swedish governments have used a particular set of instruments toobtain the objectives of economic policy. In reality, economic policies to satisfy theseobjectives have changed and some instruments are not typical Swedish. For instance,the Social Democratic government used a Keynesian strategy to secure fullemployment in the early Golden Age - labour market policies were not used seriouslyuntil the late 1950s. Furthermore, devaluations, thus, conventional measures other thandeflationary policies to reduce deficits in the current account, were made also inSweden, both in the Golden Age (1949) and regularly, in the grey period. Economicpolicy in accordance with the Swedish model could best be described generally as stateinterventionist where activity levels and employment are concerned.7

In contrast to stabilization and employment policies, the Swedish model implies apassive growth policy. It relies on the success of leading companies and industries andon measures to stabilize or improve their financial position and profitability. There isroom for infra-structural investments in the Swedish model. But these have primarilybeen intended to increase aggregate demand in a Keynesian fashion, or to strengthen“development blocks” around leading companies and industries, not to spur therenewal of Swedish manufacturing.8

It is easier to separate Sweden from other Nordic countries on economic-structuralgrounds than on the basis of political institutions, labour relations and economic policy.Norway and Finland have quite similar raw material branches, most of them dominatedby large companies, but (as Denmark) nothing like Sweden´s large, globally orientated,companies in other branches of manufacturing. Here, Sweden is more similar toSwitzerland and the Netherlands.9 However, the strong position of large transnationalcorporations in engineering in Sweden has no analogues in other small Westerncountries.

There are similarities between the Swedish model as defined in this article and Fordistaccumulation models formulated within the French regulation school, Lipietz [1988]and Boyer [1988]. The Swedish model can be seen as a regulation regime where theinterests of labour (full employment, redistribution and general welfare) are compatiblewith the interests of capital (stable profitability). My general account of the means ofthe Swedish model is consistent also with regulation theories about several possiblemodes of regulation.

7 My avoidance to be more precise about collective instruments shall not hide, that some of them areclosely connected to the Swedish model. Labour market policies but also public employment growthare such instruments. Public employment expanded more in Sweden and Denmark than in any otherOECD country in the 1960s and 1970s.

8 The passive growth policy was formed already in the beginning of the 20th century. In contrast, thestate had acted in the 19th century as an important financier of new industries in Sweden (Gårdlund[1947, p. 76]).

9 See statistics for the OECD countries, including the Nordic ones, on large-firm dependency anddirect foreign investments as a percentage of GNP (Carlson, [1979, pp. 10 and 66], Erixon [1982, p.128] and Jagrén, [1992, p. 2]).

11

But the regulation theory of the French school is too abstract to distinguish leadingindustrial sectors and conditions for accumulation in a small open country. Forinstance, collective wage bargaining is governed by the aim to stabilize demand infrench accumulation models. They neglect that large outward companies are wageleaders in a country like Sweden. There are hardly any Keynesian elements in theSwedish wage-formation model of the postwar period.

12

Chapter 3 - The Swedish Growth Engine

3.1 Raw Material Industry - A Nordic Component of the Swedish Model

The real break-through of Swedish industrialization first occurred in the 1870s. It wasbased on a surplus of agricultural workers, high foreign demand for wood products,especially from Great Britain, new mass-producing techniques, improvedcommunications, primarily through the construction of railway systems, and onabundant resources of wood, iron ore and water power. Since that time, raw materialbase industries, most of them energy-intensive and export-orientated, have been ofoutmost importance for Swedish export, industrial production and GDP. The rawmaterials´ share of Swedish export value was above 50 per cent until 1957.

I will embrace the iron and steel industries, wood product industries (including sawmills) and pulp and paper industries when the Swedish raw material sector is defined.(Iron ore is excluded by my concentration upon semi-finished and finished goods.) Theindustries relative importance did change dramatically until the Golden Age. Export ofwood products dominated Swedish export until World War I. It had decreased alreadyat the turn of the century due to harder competition from Finland and Russia. Instead,export of pulp and paper grew tremendously from the 1890s, with pauses only forWorld War I and the depressions in the early 1920s and the 1930s.10 In fact, Swedenbecame the world´s largest exporter of pulp before World War I.

In contrast to pulp and paper industries, iron and steel producers experiencedperiodical demand crises and strong foreign competition already from the 1870s. Theywere, as saw mills, seriously hit by the depressions in the early 1920s and 1930s and bysharper international competition in the 1920s. However, the Swedish export of ironand steel products rose more (both in value and volume) than export of pulp in the1930s primarily through German rearmaments.

Up to this point, I have no given Sweden a more advanced role than as a producer ofwhat is conventionally classified today as demand-fluctuating, low technologicalproducts.11 In this respect, Sweden is not different from other Nordic countries withthe exception of Denmark.12 But there is an economic-structural condition whichseparates Sweden from all its Nordic neighbours - the importance of largetransnational companies in engineering.

10 However, the depression in the 1930s was mild for Swedish paper industries.

11 However, until World War I, pulp export reduced the cyclical swings in the Swedish economy.

12 Sweden was not even a Nordic technological leader in raw material industries in the early industrialperiod, at least not in wood-product industries, cf. Gårdlund [1955, p. 248] and Sejersted [1993,chapter 3]..

13

3.2 Outward Large Companies in Engineering - An Un-Nordic Component of theSwedish Model

3.2.1 The First Innovation Wave

A cluster of innovations can be traced in Sweden from the end of the 19th century untilWorld War I. As a result, a large number of “genius companies” was established inSweden. Swedish economists have claimed, that there is no correspondence inindustrial history to the tremendous innovative strength of Sweden at the turn of thecentury and to the limited number of persons involved (Dahmén [1950, p. 366]).

Atlas, later Atlas Copco, was established already 1873 as a producer of railwayequipment, primarily carriages, for the Swedish market. However, Atlas started aproduction of pneumatic products (compressors and rock drills) in the early 20th

century. Pneumatic equipment became the company´s main product already before theend of World War I.

The telephone company (LM) Ericsson was founded by Lars Magnus Ericsson in1876. Ericsson invented a table telephone apparatus in 1884. The company developed,in collaboration with an independent company founded by Henrik Cedergren,comprehensive telephone systems for Sweden.13 The two companies merged in 1918.

The separator was an invention in the late 1870s by Gustaf de Laval which laid thefoundation of a Swedish dairy machinery industry. De Laval was one of the foundersof Separator, later Alfa-Laval, in 1883. He was also an inventor and entrepreneur inother engineering fields. The steam turbine became de Laval´s most successfulinvention beside the separator.

The electro-technical company ASEA was established in 1891 through a merger of twocompanies based on Swedish inventions, primarily engines for alternating currentdeveloped by Jonas Wenström. The company soon developed technologies related tolong-distance transfers of electricity. In 1916, ASEA required STAL, a producer ofhigh precision turbines. STAL had been founded by Birger and Fredrik Ljungström inthe 1890s.14

Inventions by Gustav Dahlén (together with Henrik von Celsing), using gas forillumination purposes, led to the formation of a company 1904 later known as AGA.Soon, AGA, with Dahlén as a manager director, got a strong position on worldmarkets, first for light-house techniques, then for gas for industrial and medical use.

13 In 1885, Stockholm had more telephone subscribers in absolute terms than any other city in theworld (Schön [1990, p. 40]).

14 STAL merged 1959 with a company based on de Laval´s steam turbine.

14

The invention of self-regulating ball bearings by Swedish technicians led to theformation of SKF (Svenska Kullagerfabriken) in 1907. In short time, the companybecame a world leading producer of ball bearings.15

It remains to explain why Sweden, but not, for instance, other Nordic regions,developed a superior innovative strength at the turn of the century.16 The reasons arenot obvious.17

Mechanisation of agriculture and of textile industry, which generated the technical baseof dairy machinery and ball bearing industries in Sweden, happened simultaneously inNorway and Denmark. Furthermore, Norway and Denmark had no less developededucational systems than Sweden at the industrial revolution.18 International contacts,established inter alia by export-orientated, raw-material industries, facilitatededucation of Swedish engineers in Germany, France, Great Britain and the U.S, buthere, Sweden did not hold an unique Nordic position. Long travel distances have beenquoted as an explanation of Sweden´s development both of long-distance energytransmission systems, and later, of heavy trucks, but in this respect, Sweden is notdifferent from Finland and Norway. Besides, the original Swedish advantage inelectrical technology was devices to use and produce electricity, not to transmitelectricity. Finally, other Nordic countries had the same financial base ofindustrialization as Sweden. Capital imports were the basis of industrialization, directlyas a financial source of infra-structural investments, indirectly as domestic saving wasreserved for industrial purposes (Gårdlund [1947, p. 163], Hansen [1972, pp. 176-177]and Hodne [1975, ch. 9]).

15 SKF integrated backward by the requirement of the steel company Hofors during World War I.

16 Other outward companies based on Swedish inventions were established at the turn of the centurysuch as Bahco 1889 (the adjustable spanner) and Esab 1904 (the welding code). Bahco becamespecialized in ventilation and air condition equipment and was finally required by Fläkt in 1987 (seefootnote 26). Esab is owned by ASEA (half of the equity stock) since 1965. The company formed ajoint company with AGA in 1987 for gas-welding operations. Swedish Match, originally SvenskaTändsticksaktiebolaget (STAB), is a company based on inventions of safety matches and machinesfor large-scale production of matches. The company was established through mergers in 1903 and1917. Succesively, Swedish Match got a diversified production including a broad range ofengineering, chemical and wood products. The company was bought and reconstructed by the raw-material producer Stora (Kopparberg) in 1988 and by Volvo in 1993.

17 Note that Sweden and Norway formed an union until 1905.

18 Until the turn of the century, higher technical education seems to have been underdeveloped inSweden in comparison to other countries including the Nordic ones. Moreover, the quality of theSwedish education was low at least until the 1870s. According to Torsten Gårdlund, the lowertechnical educational system in Sweden was underdeveloped and of low quality as well at the turn ofthe century. However, the Norwegian higher technical education might have been worser than theSwedish one. Schools of advanced technology were not started until 1910 in Norway but as early as inthe 1820s in Sweden and Denmark. (Gårdlund [1955, pp. 224-225, 230-231 and 258-259] and [1974,pp. 43-44], Hansen [1970, p. 65] and Sejersted [1993, pp. 141-142]).

15

Cultural factors, such as a strong “entrepreneurial spirit” in high income-brackets, havebeen suggested to explain Sweden´s innovative strength at the turn of the century.19

The presence of strong commercial banks from the end of the 19th century is a relatedexplanation of “the Swedish miracle”. Domestic bank credits were the prime financialsources of Swedish manufacturing until World War I. The large banks were orientatedtoward industry and, by skill, chance or clairvoyance, engaged in production withpromising market prospects and technological opportunities.20

The Wallenberg family and its bank (Stockholms Enskilda Bank) became importantfinanciers of engineering firms in Sweden. Besides, the family obtained, either from thestart or gradually, a position as a dominating owner of all genius companies, with theexception of AGA, and also of several raw material companies including Stora(Kopparberg) and the Norwegian Norsk Hydro.21 It is true that the new companiessoon became manager-led. But the Wallenberg family was an active, technical andforeign-market orientated owner and financier, engaged in strategic decisions oninvestments and selection of managers. The key position of the Wallenberg family inSwedish industrialization has no correspondence in other countries.

It is difficult to distinguish the decisive factors behind “the industrial wonder” inSweden. Comparative studies offer some clues but the explanations will varydepending on which countries Sweden is compared to. My emphasize on the industrialorientation of upper classes and the strategic role of commercial banks in Swedenfollows from my Nordic framework.22 In any case, more important from my 19 It has been argued also, that Swedish civil servants were trained and independent enough torespond to demands from a differentiated industry, Jörberg [1961, pp. 14-15]. Another argument isthat new companies in Swedish engineering were sheltered from foreign competition in their firstvulnerable phase, especially on home markets, by tariffs and patents. Separator´s posession of the“Alfa patent” 1889 -1903 is often quoted, cf. Gårdlund [1947, p. 155] and [1983, pp. 142-152 and228-234] and Jörberg [1961, p. 15].

20 Gårdlund [1947, pp. 123, 128-129, 136, 152-153 and 163], Gårdlund [1955, pp. 178-211] andGlete [1987, pp. 83-91, 111-115, 248-260 and 266-270].

21 The Wallenberg family controlled Atlas from the start, and also Atlas-Diesel, a companyestablished in 1917 through a merger between Atlas and an engine company formed and controlled bythe family. In contrast, the Wallenberg family had some minor shares only of ASEA in the beginning.But, in the first decade of the 20th century, the family played, as a financier, a decisive role in thereconstruction of the company, including in the appointment of a manager director. In the late 1920s,the family became the largest owner of ASEA and Separator (Alf-Laval) and raised its shares of thecompanies further in the 1930s and 1940s. In the early 1930s, the Wallenberg family got shares ofSTAB (Swedish Match) and (LM) Ericsson and was immediately engaged in reconstructions of thecompanies.The family became dominant owner of STAB in the late 1930s but not in Ericsson until1960 (Glete [1994, pp. 119-120]). The family led a consortium which bought ITT´s shares of Ericssoncorresponding to more than one-third of the voting-rights at annual meetings.The Wallenberg family was represented in the board of SKF already in the 1920s which facilitated thecompany´s contacts with foreign lenders. In the early 1930s, the Wallenberg family procured adominant position in the company.

22 In Norway, the upper classes were not strong or industry-orientated enough to establish largecompanies or large commercial banks. Furthermore, there were restrictions on commercial banks toissue bonds, Hodne [1975, pp. 327-329, 341-343 and 350] and Sejersted [1993, pp. 168-180].Commercial banks´ financing of large-scale industry was more important in Denmark than in Norwaybut probably not compared to Sweden. Private banks did not achieve a dominant position on the

16

perspective is that the birth and early consolidation of the Swedish genius companiescannot be related to an accumulation model where labour is a salient actor inproduction and politics or where strong and qualified domestic demand is a primedriving force.

Social Democrats did not participate in Swedish governments until 1917. Furthermore,the Swedish labour market at the turn of the century was characterized by weak tradeunions, perpetual conflicts and uncoordinated actions. Swedish workers in newcompanies had low skills and had often to be replaced or instructed by hired workersfrom other countries. Finally, Swedish innovations were not induced by demand, forinstance by higher living standards in the country anticipating world demand at highdevelopment levels. Rising welfare in Sweden was a consequence rather than a causeof the industrial wonder. It is difficult to find a better example of a supply-sideinnovation model than that in Sweden at the turn of the century.

3.2.2 The Second Innovation Wave

The growth of the new specialized industries in Sweden was tremendous. Alreadybefore World War I, they accounted for half of the production value in Swedishengineering (Gårdlund [1955, p. 86]). A new innovation frontier was opened inSwedish engineering after the war.

The second innovation wave in Sweden was not based on domestic basic inventions tothe same extent as the first wave. The term “genius” will be reserved for the companiesin the first wave. Another difference between the waves is that new companies in theinterwar period could be classified, if buses and trucks are ignored, as producers ofdurable consumption (or household) goods. The genius companies were predominantlyproducers of investment or public goods.

No other Nordic country succeeded to establish companies in mass-producingconsumer-good industries, companies which not only survived but also obtained astrong position in world markets after World War II. It is true that consumer-goods´share of export has been lower in Sweden than in continental West European countriesin the postwar period (Erixon [1982, p. 123]). But the standardized consumerindustries were important growth engines in Sweden in the Golden Age, and theirexpansion was promoted by economic policies (see chapter 6).

The Swedish firms Lux and Elektromekaniska AB introduced the first householdvacuum cleaners already before World War I. In 1919, the firms formed Electrolux, amajor transnational producer of vacuum cleaners in the Golden Age. Electrolux began

Danish credit market around the turn of the century. In addition, they were not as (export-) industryorientated or active in the formation of new industries as their Swedish correspondences. Besides, thedominating credit institutions in Denmark, saving banks and credit associations, were primarilyengaged in the financing of farming and construction, not of industry, Hansen [1970, pp. 28-67].Information about Finland is incomplete, but it seems that commercial banks did offer short-termcredits only to industry about the turn of the century (Harmaja [1933, p. 43]).

17

to produce refrigerators in 1925 after a path-breaking invention by two young Swedishtechnicians.

The first motor vehicle firm in Sweden was established in the early 1890s. A merger in1911 between two independent car producers led to the foundation of Scania-Vabis,specialized in trucks and buses in the 1920s. Volvo was established in 1926 as aproducer of private passenger cars and, somewhat later, of trucks and buses. Thecompany grew by take-overs in the 1930s and 1940s to enlarge its motor vehicleassortment and to integrate backward inter alia to motor industries.23

The other Swedish automobile company (beside Volvo) in the postwar period, Saab,was founded in 1937. Saab was first engaged in production of aeroplanes only - thecompany´s car production did not get into its stride until after World War II.

The explanations of the first innovation wave are probably valid also where productrenewals in Sweden in the early interwar period are concerned. Once again, industrialcapitalists played an active role as financiers and reconstructors. An outwardentrepreneur, Axel Wenner-Gren, initiated the development and sale of householdvaccuum cleaners and also the foundation of Electrolux where he became major ownerand manager director. Saab was formed by Electrolux and by another company(Bofors) controlled by Wenner- Gren. The Wallenberg family did early take the controlover Saab and, successively, over other companies in the second innovation wave.Only one of the large second-wave companies remained outside the Wallenberg spherein the Golden Age - Volvo.24

It seems also that Swedish engineering experienced a virtuous circle in the interwarperiod. Technological knowledges and profits emanating from the first wave ofproduct innovations led both to secondary innovations and to new industries.25 Forinstance, the original electrical development block was extended, either by new firmsor by established firms as ASEA and AGA, to the production of trains, trams andhousehold products such as stoves, vacuum cleaners and radio sets.26

23 Volvo´s requirement of the engine company Pentaverken (1935), today Volvo Skördeverken andVolvo Penta, and the machine company Köpings Mekaniska Verkstad (1942) were both examples ofVolvo´s ambition to broaden its motor vehicle programme and to integrate backward. Volvo did alsorequire Svensk Flygmotor in 1941, a producer of aeroplane engines founded in 1930.

24 The Wallenberg family gained control over Scania-Vabis during World War I. The company gotinto liquidation in 1921. The Wallenberg family was not engaged in product developments and thechange of product composition in Scania-Vabis in the 1920s when, for instance, the company gave upits car production. (Scania-Vabis´ production of buses and heavy trucks became profitable already inthe late 1920s.) However, the family took an active part in the standardization (and rationalization) ofthe company from the late 1930s (Giertz [1991, pp. 123-170]).The Wallenberg family became major owner of Electrolux in the latter part of the 1950s and the early1960s through its possession of Alfa-Laval and ASEA.

25 Important secondary product innovations (although primarily by foreign firms) did also occur inraw material industries in the interwar period, see Dahmén [1950, pp. 124-125, 137-148 and 157-161].

26 There are significant examples of new Swedish companies within the electrical development blockin the interwar period. Electro-Helios, a producer of electrical household products, was established in

18

The virtuous circles in Swedish engineering were not only technical and financial butalso personal. The major share of entrepreneurs in engineering in the interwar periodwas recruited from skilled labour (Dahmén [1950, pp. 227-228 and 386-388]). Thebirth of the companies in the second innovation wave which later became growthengines in Sweden can be related to established companies and owners inengineering.27

There were clear evidences of industrial “maturity” in Sweden in the 1920s anddefinitely in the 1930s. Some companies (especially from the first innovation wave)and owners (the Wallenberg family) then procured dominant positions in Swedishmanufacturing, positions which they still hold today. From the 1930s, the success ofSwedish industry largely depended on the ability of managers and main owners ofestablished companies to adjust to new market opportunities and threats. For instance,the formation and development of new industries in Swedish manufacturing mainlybecame an affair of some large companies and capitalists in engineering. Thus, it hadbeen legitimate to use a static notion such as “the Swedish growth engine” already inthe 1930s.

3.2.3 The Early Foreign Orientation of Swedish Companies

“Outwardness”, or more precisely, “early outwardness”, will be seen as a preconditionfor the success of the large Swedish companies in engineering in the Golden Age.Thus, foreign orientation is considered as a precondition for the Swedish model.

The companies in the first innovation wave pursued already in their early phases ofdevelopment a strategy of conquering foreign markets. They soon worked toovercome the limitations of the Swedish markets by foreign sales and also to useeconomies to scale to meet foreign competition. Already in the 1890s, Swedishengineering consisted of two groups of firms, one of large export orientatedcorporations, another of small and medium sized, home-market orientated and non-specialized firms (Dahmén [1950, pp. 225-229] and Jörberg [1961, pp. 77-79]).

1919. Fläkt (earlier Svenska Fläktfabriken) was formed in 1920, an early outward company whichbecame specialized in devices for heating, freezing, ventilation and air conditioning. Today thecompany is world leading in smoke-gas purification, energy regaining and waste treatment. Luxor,established in 1918, became a producer of radio sets, gramophones and loud-speakers. The companybecame one of the few large ones in Swedish computer industry in the 1970s (producing computersmainly for industrial purposes) but was finally required by the Finnish company Nokia in 1984.

27 The birth of Volvo could both personally, financially and technically be traced to SKF. In fact,Volvo was a part of SKF until 1935. The vacuum-cleaner company Lux was a subsidiary to AGA. Thefounder of Electrolux, Axel Wenner-Gren, was employed by Separator (Alfa-Laval) before he initiatedthe Swedish vacuum cleaner industry. As already mentioned, Saab was established by Electrolux andby its main owner, Wenner-Gren. Finally, Saabs early expansion was partly financed by Wallenberg´scompany Aktiebolaget Svenska Järnverkstäderna. The formation of Scania-Vabis cannot directly betraced to established companies and owners. However, SKF played an important technological role inScania before the merger with Vabis.

19

ASEA, however, has traditionally been more home-market orientated than othercompanies in the first innovation wave.

Thus, domestic-demand restraints explain why Swedish genius companies becameoutward so early. But companies in the second innovation wave did not meet earlydomestic market-restraints and they were often foreign-orientated as well.28 “Earlyoutwardness” of Swedish companies was reinforced by political factors. A sharpincrease in the export share of total sales in the 1920s, especially for genius companies,is largely explained by enlarged international trade and restrictive Swedish monetarypolicies. (Analogously, the decline in export shares in the 1930s is largely explained byprotectionism and expansive monetary policies.) In the case of (LM) Ericsson, foreignorientation was hasten by governmental regulations not allowing the company to sellon domestic markets (Sölvell et al [1991, p. 117]).

Early export orientation of engineering companies is also explained by behaviourfactors such as the existence of extrovert Swedish managers and major owners. Theimportance of “outwardness”, irrespective of domestic demand-restraints and politicalconditions, is best shown by a company in the second innovation wave, Electrolux.The company´s early foreign orientation is striking in comparison to both foreign andSwedish competitors. Another, partly related, factor behind the early foreignorientation of Swedish companies, is the existence of international net-works fortechnicians, capital owners and firm executives. The net-works had been founded bythe raw-material industries but also by companies in the first innovation wave.

The emphasize on informal net-works when the outward strategy of Swedishcompanies in engineering is to be explained illustrates the complexity of causal analysisin social science. A “voluntary” phenomenon, personal contacts across nationalborders, is explained by a structural condition, the existence of outward raw -materialcompanies in Swedish industry.

There is another side of “outwardness” than export orientation - foreign production.Swedish engineering companies in the first innovation wave did early establishsubsidiaries abroad. The main explanation is “structural”. Foreign demand for products 28 Foreign sale departments for vacuum cleaners were established by Electrolux from 1919.Electrolux´ first production plant abroad for vacuum cleaners was built in 1925 (in Berlin) and forrefrigerators in 1927 (Luton), Laxén-Payro & Odhnoff [1985, p. 7]. Volvo was also export orientatedfrom the start. The company established a department for foreign sales of cars and trucks as early asin 1928 (in Helsinki, Finland). In the late 1930s, export accounted for 20 percent of Volvo´s total sale(Plate [1985, pp. 18-24]). However, Volvo did not start a foreign production until 1958 (in Finland).The case of Scania-Vabis´ is both a confirmation and a rejection of a theory of “early foreignorientation”. The company made export businesses from the start and established assembly plants inDenmark, Norway and Russia as early as in 1913-1914 (Giertz [1991, pp.98-99]). An interruption inScania-Vabis´ export occurred from the late 1910th to the end of the 1930s . In this period, thecompany had public authorities in Sweden as its only customers. The export recovery of Scania-Vabisat the end of the 1930s was rather slow. Export accounted only for 13 percent of total salesimmediately after World War II. It was not raised considerably until the mid 1950s (Giertz [1991, p.237]) . Scania-Vabis first plant abroad after the break in foreign orientation was established as late as1957, in Sao Paulo, Brazil. As Scania-Vabis, Saab did not increase its export seriously until the mid1950s. The company´s foreign production of private passenger cars began in the early 1970s, inNystad, Finland.

20

from ASEA, LM Ericsson, Alfa-Laval, SKF and AGA must, by technical andmarketing reasons, or by conveyance, be met by production abroad (Carlson [1979,pp. 43-47]). The conclusion that foreign markets must by necessity be supplied bysubsidiaries abroad is valid also for refrigerators, thus for some products in the secondinnovation wave,. However, a strong Swedish krona (SEK) might have contributed toa sharp rise in Swedish companies´ production abroad in the 1920s.

Early foreign sales was not always synonymous to early foreign production for thecompanies in the first innovation wave. Atlas Copco did not establish a foreign unit forpneumatic products until the late 1930s (in Great Britain).

Swedish companies in the first and second innovation wave reject traditional tradetheories about engineering saying that small countries will specialize in market nichesand small scale production. They also contradict the product-cycle theory saying thatexport, foreign production and standardization are typical of later development phases.Furthermore, they do not confirm John H. Dunning´s theory that foreign directinvestments in the early stage of a country´s development prevent upgrading ofdomestic technological capabilities and reduce the competitive advantage of both theinvesting firms and the country itself (Dunning [1993, p. 76]).29 Finally, the early“outwardness” of Swedish genius companies seems as weak evidence of a theorystating that qualified home market demand is a spring-board to internationalcompetitiveness (cf. Porter ([1990, pp. 86-99]).

However, the objection to the last theory must be modulated. Qualified demand fromraw material and related industries (such as mining) and from political authoritiescontributed to the foreign success of Swedish engineering. Thus, other domestic agentsthan engineering companies must be accounted for when the Swedish innovationmodel is defined (see next chapter).

It remains to argue for that raw material industries and large engineering companiesheld strategic positions in the Swedish economy in the Golden Age. My emphasize islaid on the Swedish innovation system. As my attempt is to distinguish unique featuresof Sweden, I have no intention to formulate a general innovation theory, not even forsmall open economies in general.

29 It seems, however, that Dunning leans to the conclusion that countries like Sweden are not hit bythis criticism of an early outward strategy (Dunning [1993, pp. 116-117]).

21

Chapter 4 - In What Way Are the Leading Industries Leading?

The raw material industries and large engineering companies were growth engines inSweden in the Golden Age by the following reasons:

- They accounted for a large share of production and employmentin manufacturing.

- Their export and investment had a strong influence on productionand employment of other Swedish firms.

- They were wage leaders.

- They were the most important producers and transmitters of new knowledge.

4.1 Shares of Production and Employment in Swedish Manufacturing

The raw material sector accounted for more than 20 per cent of total value added inSwedish manufacturing from the early 1950s to the mid 1970s - see table 6. Thesector´s share of total employment in manufacturing was generally smaller because ofhigh capital intensity (table 6). Both shares were remarkably stable in the Golden Age.

The large companies in engineering accounted for a large and growing portion ofemployment in Swedish manufacturing in the Golden Age. The list of Sweden´s fifteenlargest companies, where employment in Sweden is concerned, was dominated byengineering and raw material industries (with ASEA as the largest) in 1945. At the endof the Golden Age, the list was dominated by engineering companies only. The growthof Saab, Volvo and Electrolux in the Golden Age is striking. They all reached aposition as one of the five largest companies in Swedish manufacturing from a positionoutside the list of Sweden´s fifteen largest companies at the end of the war.30

30 Unfortunately, there are only postwar figures of the 15 largest companies in Swedish manufacturingwhere employment in Sweden is concerned for 1945, 1983 and 1990, thus not for the late Golden Age(Jagrén ([1992, p. 9-11]). However, Volvo and Electrolux reached the top five (and Saab(-Scania) thetop eight) in 1973 if total employment is considered that is employment in Sweden and in subsidiariesabroad (The 1000 Largest Companies in Sweden [1974, p. 109]).Elecrolux would have entered the list of the fifteen largest companies only a few years after 1945 or ifemployment abroad had been considered. The company was one of the five largest Swedish companies

22

Saab, Volvo and Electrolux showed high internal growth in the Golden Age but theydid also expand by mergers (mainly Saab) and take-overs. The aim of requiring otherfirms was to either become national champions (mainly Electrolux), get broaderproduct assortments (mainly Electrolux and Volvo) or integrate backward (mainlyVolvo).31

The large engineering companies´ larger shares of Swedish employment in the lateGolden Age are shown by table 8. Alfa-Laval, ASEA, Atlas Copco, Electrolux,Ericsson, Saab(-Scania), SKF and Volvo accounted together for 13 per cent of totalemployment in Swedish manufacturing 1966. The share increased to 15 per cent in1970 and to more than 17 per cent in 1973.32

4.2 Net-Works and Multipliers

An extensive net-work of supporting industries has emerged around the raw materialindustries and large engineering companies in Sweden. Swedish manufacturing wascharacterized in the Golden Age by a significant number of supplying firms stronglydependent on orders from the large companies. This dual economy can bedistinguished, as already noted, in the 19th century.33

In addition, mutual orders linked the large raw material and engineering companies toeach other in the Golden Age. Investments by capital- and energy-intensive rawmaterial firms favoured engineering companies such as Volvo, ASEA, Alfa-Laval andSKF (Sölvell et al [1991, pp. 81-82 and 114-115]). Inversely, the iron and steelindustries in Sweden were favoured by investments in engineering (including shipbuilding).

Thus, the raw material industries and large engineering companies were important forSweden in the Golden Age not only through their large share of production andemployment in manufacturing. They also acted as spiders in strategic industrial net-

already in the mid 1930s when employment in foreign subsidiaries are accounted for (Wibe [1976, pp.22-23]).

31 Volvo´s earlier strategy to extend its motor vehicle program and integrate backward continued inthe Golden Age. The company required a producer of agricultural vehicles, Bolinder-Munktell, in1950. Volvo´s backward integration strategy culminated with the take-over of a truck-cabin producer(given the name of Volvo Umeverken) in 1964 and a steel-pressing company (Olofström AB) fromAlfa-Laval in 1969.

32 Figures for AGA is also included in table 8. Some engineering companies such as Fläkt, Bahco andEsab are excluded since they were not large enough or owned by other large companies in the GoldenAge. Swedish Match has been excluded on statistical grounds.

33 The supplying firms, however, became more specialized in the interwar period (Dahmén [1950, pp.239-241]).

23

works. Besides, the industries and companies under discussion had strong demandmultiplier effects on other parts of the Swedish economy than supplying industries. Inthese respects, export was a more important activity than investment.

The raw material industries accounted for more than 60 percent of Swedish exportvalue (for goods) immediately after World War II and in the early 1950s - see table 7.Their share of Swedish export then decreased steadily but was still 38 per cent in 1973.The sector´s share of Swedish net exports was even larger since it predominantly usesdomestic inputs. As a contrast, in Norway, the influence of raw material export on therest of the economy is weaker because of high import shares of production costs.

The eight companies in engineering mentioned above accounted for 19 per cent of theSwedish export value in 1966 - see table 9. The share increased to 22 per cent 1970and to 24 per cent 1973. The increases were stronger than for engineering products asa whole primarily because of Volvo´s and Saab(-Scania)´s export successes.34

Export became, together with housing demand and investment in construction, themain demand driving force in Sweden in the Golden Age. Surpluses in the currentaccount were not counter-balanced by revaluations of the SEK (see the Bretton Woodsystem) and they also created room for expansive fiscal and monetary policies.35 As aconsequence, export from the leading companies had a strong impact on Swedish GDPin the Golden Age.

4.2 Wage Leadership

The decisive influence of export on Swedish GDP has given the big companies in basicindustries and, particularly, in engineering a wage-leading role. Other companies areforced to follow since they are anxious to keep key employees and also restricted bywage earners´ care for relative wages. In the Golden Age, the relative wage preferencewas strengthened by solidaristic wage policy and also fairly easy to satisfy because of agrowing scarcity of labour.

Wage leadership by the export sector is emphasized in the Scandinavien wage modelformulated in the 1960s (Edgren, Faxén & Odhner [1970]). Here, central wagenegotiations are not independent determinants but rather “collective replies” to wageincreases for (key) employees in the exposed sector. There are no indications that theScandinavien model became less valid for Sweden in the 1970s and 1980s (Holmlund[1990] and Warginger [(1992]).

Wage leadership is important not only inflation but also for productivity and growth.Wage increases in large export companies lead to a cost pressure on other firms,

34 The export figures for the raw material industries and the large engineering companies have notbeen added up since some engineering companies are also engaged in steel industry.

35 Current-account deficits did emerge in the Golden Age since economic policies immediately afterthe war (1947-1949) and in the economic peaks of the late Golden Age were too expansive. But thedeficits were rather small or of short duration.

24

including new ones. The cost pressure will result in higher productivity through eitherrationalizations in suffering firms or structural change. (The decline of less-effectivefirms and industries will release resources for the expansion of other firms andindustries.)36

4.4 The Production and Diffusion of Knowledge by the Large Engineering Firms

The Swedish innovation system in the Golden Age is a mixture between a fundamentalsystem, settled already in the interwar period, and a “regulated” (collective) one whichonly partially could be distinguished before World War II. The two systems arecomplementary and reinforcing. In both, large established companies in engineering arethe main producers and transmitters of new knowledge. The difference between thesystems is that the first is based on early patterns of specialization andinternationalization while the second one incorporates the institutions and means of thepostwar policy model as defined in chapter 2.

The Swedish innovation model assumes that large firms produce a larger share of newideas than other domestic firms, single inventors and universities. This assumption isconfirmed by studies of major inventions during the Golden Age even when large firmsare compared to all other inventors (Mc Queen & Wallmark [1983, pp. 48-49]) andCarlsson et al [1979, pp. 141-142]). I will also assume, that the inventive superiority oflarge firms primarily reflects the superiority of large companies in engineering. Thetheory has not been tested rigorously, but there are many indications of its validity (McQueen & Wallmark [1983, pp. 11-16, 28 and 42-49]). In addition, the spill-overeffects of inventions in engineering are probably stronger than in other industries, forinstance pharmaceutical industry.

An inventive superiority of large Swedish companies in engineering might be explainedby their financial strength, facilitating R&D expenditures, their long market experience,producing new ideas by feedbacks from customers and competitors, and by scaleadvantages in R&D.

The inventive strength of Swedish engineering companies laid the ground for a stronginnovative ability in the Golden Age. The companies´ inventive sovereignty might havebeen amplified by superior sale and development capabilities reflecting long marketexperiences or rich financial endowments allowing high expenditures on R&D andmarketing. Their innovative vigour could also have been due to scale advantages inmarketing and product improvements.37 In any case, innovative superiority is one

36 In the Scandinavien wage model, productivity is exogenously determined.

37 The inventive rather than the commercial superiority of large firms in Sweden is emphasized in aSwedish study (Mc Queen & Wallmark [1983, pp. 45-49]). Moreover, the size of R&D investmentsseemed to be a stronger comparative advantage of large engineering firms than R&D efficiency. Thelarge engineering firms´ share of Swedish major innovations in the Golden Age (1945-1973) was 38per cent which is approximately equal to their share of value added in manufacturing. The formershare is 43 per cent if pharmaceutical firms are excluded. (My estimations are based on Mc Queen &

25

explanation why large established companies warranted a large proportion of newindustries in Sweden in the Golden Age.

The innovative sovereignty of large engineering companies in Sweden is, to a largeextent, based on extensive R&D activities and expenditures. The corporate sectoraccounts for a dominant share of R & D activities in Sweden. In addition, R & Dinvestments are concentrated to the large companies in engineering and topharmaceutical companies. (However, a similar concentration of R&D investments tothe corporate sector and the large companies in manufacturing can be found in otherOECD countries.) Moreover, approximately 90 per cent of total R & D expendituresin Swedish manufacturing has been financed by the companies themselves in thepostwar period.38

4.4.1 The Fundamental Innovation System

When the fundamental innovation system in Sweden is distinguished, I will notconsider the early innovation process in engineering. Thus, I will ignore themechanisms behind the establishment and early consolidation of firms in the first andsecond innovation wave. My focus is upon innovations in established companies inSwedish engineering, thus, upon secondary innovations or the shaping rather thanmaking of leading companies in Sweden. In this perspective, the relationships betweenthe large Swedish companies in engineering are of less interest. In the postwar period,there are few examples of secondary innovations in leading companies which can beattributed to qualified orders from or collaborations with other leading companies inengineering.39 For instance, the technological links between SKF and the Swedish carproducers have been weak. In contrast, SKF was a midwife at the birth of the Swedishcar industry. Basic innovations in Scania(-Vabis) and Volvo can be related to productdevelopments in SKF.

There are two fundamental impulses to innovations in large engineering companies inSweden. The first impulse emerge through their “outwardness”, the other throughnatural conditions or the Swedish raw material industries.

The large Swedish companies in engineering assimilate and develop new ideas throughtheir international connections. Demanding foreign customers and tough foreigncompetitors encourage product differentiation and the introduction of newtechnologies in the Swedish companies. Innovations by foreign competitors are Wallmark [1983, table 3]). There was, on the other hand, a strong correlation between R&Dexpenditures and major innovations in Swedish manufacturing in the Golden Age (Mc Queen &Wallmark [1983, pp. 43-44]).

38 Statistics Sweden [1989, p. 21] and [1992].

39 Alfa-Laval and Electrolux started some commercial and technical collaboration in the mid 1950s.SKF and ASEA co-operated in the mid 1960s to develop flexible steel producing methods. Saab, (LM)Ericsson and ASEA collaborated in space technology in the mid 1960s. Furthermore, Alfa-Laval andVolvo had some technical collaboration until 1969 as Alfa-Laval was a supplier of car bodies toVolvo.

26

imitated or simply appropriated through take-overs. Swedish transnational companiesdevelop and acquire new ideas also by joint ventures together with foreign firms andthe use of foreign suppliers.

It is misleading, however, to consider foreign orientation as the only fundamental forcebehind (secondary) innovations in Swedish engineering. For instance, Scania-Vabismade product innovations (and radical organizational changes) in the interwar periodalmost without any export at all until the mid 1930s.

Innovations in Swedish companies in engineering have been promoted by hard naturalconditions. Product developments in electro-technical industry (ASEA) and in motorvehicle industry (Scania-Vabis and Volvo) were stimulated by long distances andsevere climate. (The “regulated” innovation system has also been important for thesecompanies.) Furthermore, Swedish mining conditions influenced Atlas Copco todevelop light pneumatic drills with pusher legs, the so called Swedish method, whichbecame a world-market success in the Golden Age (Janelid [1974, pp. 271, 287 and292-293]). In addition, innovations in engineering industries can be related to otherleading industries in Sweden. Product developments in Volvo, Saab, Atlas Copco,ASEA and SKF have been stimulated by qualified orders from and technicalcollaboration with Swedish raw material producers.

But I will consider foreign orientation as the most fundamental force behind productdevelopment in Swedish engineering in the Golden Age. Natural conditions and rawmaterial industries were probably more important before the Golden Age.

The ideas produced or acquired by the large Swedish transnational companies aretransmitted to their units in Sweden and to other Swedish firms. There are manypossible mechanisms of transmission. For instance, employees in large engineeringcompanies establish new firms on the basis of ideas acquired in their formeremployment. Moreover, new management theories are spread by central employerorganizations dominated (at least from the early 1930s) by the large companies.

Due to firm concentration, new knowledges in the leading companies in engineeringare spread to complementary, not competing, firms and to other industries, inter aliasupplying ones. Michael Porter has argued that many domestic producers in anindustry are necessary to create a dynamic environment. In fact, he mentions the rivalrybetween Saab and Volvo (both passenger cars and trucks) as a good example (Porter[1990, pp. 117-122]). But here, Porter has found the exception of variety rather thanthe rule of solitude in Swedish engineering. The large companies are, with someexception, the only producers in their fields.40 They illustrate that passionate struggles

40 ASEA, Separator and SKF had Swedish competitors for main products but they disappearedthrough take-overs and closures in the interwar period. Atlas Copco became the only Swedishproducer of pneumatic equipment by the requirement of two producers of small compressors in 1942and 1948 respectively. In addition, Scania-Vabis and Volvo have been the only Swedish producers oftrucks and buses since the mid 1930s when a competitor closed down. Electrolux became the soleSwedish producer of vacuum cleaners in 1934 and, although much later, of refrigerators, freezers andelectric cookers by take-overs of Electro-Helios in 1962, earlier controlled by ASEA, and Husqvarnain 1978. At the end of the 1970s, only one domestic white-good competitor to Electrolux remained,Cylinda (owned by ASEA until 1988), a producer of washing machines and dish washers. In all cases,

27

between domestic rivals can efficiently be replaced by competition between nationalchampions, particularly if some of them have been formed, as the Swedish ones, by anoutward strategy. The Porter model assumes that innovations are promoted bydomestic competition and then spread by intra-industrial interactions and labourmobility. However, in Sweden, new ideas have traditionally been diffused by nationalchampions to firms in other industries, including subcontracting ones.

The raw material industries are not only inspirers of new ideas in engineering in theSwedish innovation system. They are also receivers of new ideas, developed ortransmitted by the large engineering companies. In adition, the use and development ofnew technologies and products in raw-material industries are speeded up by co-operation with and qualified orders from the large engineering companies (Janelid[1974, pp. 278-279 and 306] and Schön [1990, pp. 40-41, 59 and 81-82]).

The relations in the fundamental innovation system are one-sided if the largecompanies in engineering and basic industries are added up. Knowledge produced orassimilated by these companies is spread to other companies in Sweden but not theother way round. Here, some obvious cases where small and medium sized firms inengineering have played a innovative role in Sweden are ignored. The fundamentalinnovation model does not exclude that new ideas are developed by small domesticfirms. However, the model assumes that new ideas are commercialized by establishedlarge companies by appropriation, either of the ideas or of the small firms themselves.Established companies´ requirement of dynamic new firms is one reason behind the fewamendments to the group of large companies in Sweden, a pattern already fixed in theinterwar period.41

4.4.2 The Regulated Innovation System

In the “regulated” Swedish innovation model, governments, both central and local, andpublic agencies are main actors together with established large companies inengineering. Innovations in the latter companies were brought forth by the followingdevices in the Golden Age;

• Regulations, primarily related to housing standards, energy consumption and investment, safety and environment (including work environment).

• Qualified public orders.

• Technical co-operation with public authorities.

• Economic policy including tax policy. however, the noticed companies were branch-leading even before the time of “domestic monopoly” (or“domestic oligopoly” as in the case of Swedish car industry).

41 For instance, ASEA required the major share of Fläkt in 1933 and incorporated Electro-Helios in1936.

28

It can be objected that innovations in other Swedish (engineering) companies than thelarge ones have been promoted in the postwar period by public orders, technical co-operation with public agencies and regulations. Another possible objection is thatdomestic producers have been favoured by government purchases and economicpolicies in other countries as well. Thus, the Swedish “regulated” model of innovationis not unique.

It is true that small and medium-sized firms have participated in public systems ofinnovation in Sweden in the postwar period. But the co-operative and monopsoniccharacter of public demand favoured large suppliers. Furthermore, there are severalreasons why the regulated model was more valid for Sweden than for most otherOECD countries in the Golden Age apart from the obvious one, that public sectorgrowth (whatever measure is chosen) was stronger in Sweden than in other OECDcountries (with the exception of Denmark) in the 1960s and the 1970s.

First, Swedish large companies in engineering are specialized in investment goods, ormore exactly in public (infra-structural) investment goods. ASEA (electro-technicalequipment) and Ericsson (tele-communications) are the best examples.42 Second, thelarge Swedish companies in engineering were strongly favoured by militaryrearmaments in Sweden after World War II. As a result, a modern aircraft industry(including a civil division) was built up by Saab and jet engines were developed(independently) by Volvo and ASEA in close co-operation with Swedish defenceauthorities. Furthermore, strong demand from defence during and after World War IIinduced Scania-Vabis (later SAAB-Scania) and Volvo to develop durable cars, heavytrucks and buses. It also led to the development of radio-communication systems byEricsson.

Third, the “regulated” innovation model was exceptionally important in Sweden in theGolden Age since welfare policies stimulated demand of standardized consumer goods.For instance, the strong priority of decent housing in Sweden favoured large producersof household appliances such as Electrolux. The so called Million Programme 1965-1974, that is, the Social Democratic attempt to put an end to low -quality living andhousing shortage, was extremely important for domestic white-good producers. White-good consumption grew significantly more than total private consumption in the lateGolden Age (cf. table 13). However, the Million Programme´s effects on productrenewal in white-good industry were probably limited.

42 A collaboration between ASEA and the Swedish State Power Board (Vattenfallsstyrelsen) resultedin a high-tension transmission system adapted for the water power station Harsprånget built in 1952.Similar transmission techniques were later used in other countries. A similar collaboration withpublic authorities induced ASEA to start a production of nuclear power in the 1950s. ASEA becameboth a technological leader and a supplier of nuclear power in Sweden in the 1960s. In fact, thecompany built nine of twelve nuclear plants in Sweden (Glete [1987, p. 278] and Schön [1990, pp. 82-83]). A technical collaboration between Ericsson and the Swedish Telecommunication Administration(Televerket) from the early 1970s led in the middle of the decade to the introduction of a pathbreakingdigital switching technology, the so called AXE system (Sölvell et al [1991, pp. 117-118]).

29

Innovations in large engineering companies were stimulated also by stabilization andtax policies in the Golden Age. It is assumed here that innovations are promoted byhigh domestic demand or facilitated by internal saving. Import regulations are alsoincluded in my notion of stabilization policies. It is assumed here, that productdevelopments are promoted by the escape of superior foreign competition, at least fora limited period.

30

Chapter 5 - The Performance of the Swedish Model in the Golden Age

5.1 Threats, Opportunities and Reactions I - The Large Engineering Companies

Large Swedish companies in engineering experienced a tremendous success in worldmarkets in the Golden Age. Established companies such as Alfa-Laval, Atlas Copco,Ericsson, SKF, AGA and ASEA succeeded not only to preserve but also to strengthentheir technological and commercial positions, Carlsson et al [1979, pp. 138-154].Scania-Vabis and Volvo were among the world´s largest producers of buses and heavytrucks in the Golden Age. They became technological leaders by developments of theturbo engine already in the early 1950s. Where private passenger cars are concerned,the expansion of Volvo and Saab was not curbed by the fact that they were smallproducers in standardized markets. For instance, the turbo engine for private passengercars was mainly an inovation by Saab in the first half of the 1970s.

Electrolux strengthened its position as a large producer of vacuum cleaners on theworld market in the Golden Age. In addition, the company strove successfully toextend its white-good programme from refrigerators to washing-machines,refrigerators, freezers, electrical stoves and dish washers and for a position as adomestic monopolist. Electrolux did not become a world leading producer in theGolden Age but the company´s strategy to “buy market shares” was shaped bydomestic take-overs. The strategy was brought to perfection in the 1970s and the1980s when Electrolux gained a strong oligopolistic position on world markets forwhite products.

Swedish companies in engineering were favoured in the Golden Age by the rebuildingof Europe and liberalization of trade inter alia through GATT and EFTA agreementsfrom the late 1950s. Swedish engineering had a favourable regional and productcomposition of its export at least until the mid 1960s.43 In addition, Swedishcompanies were favoured by the fact that Sweden was one of the few Europeancountries that had not participated in World War II.

But favourable external conditions are no guarantee of business success. Besides,market conditions were not unambiguously favourable for Swedish engineering in theGolden Age. First, trade extensions and a rapid diffusion of new knowledge resulted ina hard competitive pressure on large Swedish companies in engineering already in the1950s. Swedish companies competed mainly with Anglo-Sachian companies which hadbeen less hit by the war. (Here, Saab, Scania-Vabis and Volvo are the clearestexceptions.) Besides, Swedish companies were large in relation to other Swedishcompanies but not in relation to foreign companies. Scale advantages of foreigncompetitors were often stronger in marketing, R & D, distribution and finance than in 43 Export demand from the EFTA countries was high in the first half of the 1960s. However, theregional composition of Swedish export of engineering products became unfavourable from the mid1960s (Mjøset et al [1997]).

31

production.44 Moreover, researchers disagree whether large-scale productiontechnologies became more or less important in the Golden Age. In any case, therewere strong incitements at least for Swedish suppliers of consumer durables to adaptlarge-scale technologies as markets were enlarged, standardized and penetrated byforeign companies with absolute scale advantages.

Second, leading Swedish companies were hit by market saturation in the Golden Age.A theory that export market growth of the leading Swedish companies became lowerin the late Golden Age is not unambiguously confirmed by empirical research.45 But itis obvious that many domestic markets were exhausted, especially in the 1960s.46

Large Swedish companies, however, had the qualifications to meet the challenges inthe Golden Age. They were often more accustomed to free trade and foreigncompetition than rivals from large Western countries. Trade restriction in the 1930sand 1940s were parentheses for Swedish companies which had been developed on thepremises of foreign sales. Even ASEA, the most home-market orientated companyfrom the first innovation wave, was more outward (and specialized) during the GoldenAge than foreign competitors.

“Early outwardness” facilitated Swedish companies´ take-over of foreign companiesand their management of subsidiaries outside Sweden. There was a sharp increase indirect investments abroad by Swedish companies in the 1950s and the 1960s. Thecompanies expanded predominantly abroad by take-overs (SOU 1982:27 [1982, pp.69-72 and 113-115]).

In addition, the probability to become a Swedish champion in the Golden Age wasraised if the companies had a long experience of export and foreign production.Electrolux became a national champion for white goods, primarily by take-overs, in theGolden Age. The main difference between Electrolux and its Swedish challengers wasthat Electrolux was much more outward. Analogously, Electrolux´ foreign experiencesis an important explanation of why the company got a world leading position for whitegoods (mainly by take-overs) in the 1970s and the 1980s.

Leading Swedish companies did respond to harder international competition not onlyby take-overs but also by mergers. A merger between Saab and Scania-Vabis in 1969

44 However, Atlas Copco became the world´s second largest manufacturer of pneumatic products inthe early 1960s (Ramström [1974, p. 344]).

45 Mjøset et al [1997] do not show that the product composition effect of engineering on Swedishexport market shares was unfavourable from the mid 1960s particularly not if ships are excluded (seealso Hultén [1988, pp. 194-195]). However, a more detailed study has added refrigerators to the groupof engineering products with a low export growth (Hultén [1988, p. 220]). Beside, a study of the largeengineering companies in Sweden confirmed a theory that the most outward companies (companieswith the largest foreign production) experienced the lowest export growth. The study covers not onlythe Golden Age but also the 1975-1983 period (Hultén [1988, pp. 196-202]).

46 For instance, growth of private consumption expenditures on new cars (constant prices) decreasedsignificantly in Sweden in the 1960s. In fact, it became lower than growth of total privateconsumption expenditures in the 1963-1973 period - see table 13.

32

reflected a need to use economies of scale in production, R & D, marketing andadministration but also to stabilize demand for the two companies. (However, thehopes of scale advantages were largely dashed.)

Large Swedish companies raised their R&D expenditures significantly in the 1960s andearly 1970s to meet the challenges from foreign competitors. As a consequence, civil R& D spending as a percentage of GDP increased more in Sweden than in most otherOECD countries (Fagerberg [1987, p. 122]). R & D efforts were probably stronger inlarge Swedish companies in engineering than in competing foreign companies.47

R & D expenditures were important for the success of Swedish companies in theGolden Age. But the overwhelming majority of innovations was made abroad. LargeSwedish companies were fast to use both flexible production technologies, industrialrobots and new rationalization instruments such as the MTM system.48 They had beenequally efficient in the interwar period to catch up large-scale innovations from the U.Ssuch as the assembly line and scientific management (Taylorism).49 The fast Swedish“catch-up” after World War II is largely explained by international net-worksestablished by the large companies in engineering.

The incitements in large Swedish companies to rationalize were raised in the 1960s dueto harder external pressure and “aggressive” market strategies to conquer worldmarkets (see Electrolux in particular). Organizational changes in Swedish engineeringwere still mostly of a “Fordist” nature leading to work specialization rather than towork differentiation (Carlsson et al [1979, p. 129-130]). Cost-reducing programmeswere accomplished not only for single plants. ASEA and Electrolux in particularconcentrated production by plant specialization.

Furthermore, the large Swedish companies in engineering reacted in the 1960s on tightinternational competition by specialization on “core products” - often their basicinnovation - and economies of scope within the core.50 An important part of thestrategy was to sell out or close divisions with weak commercial or technological links

47 This is indicated by figures on R & D expenditures as a ratio of sales values in 20 manufacturingindustries in Sweden and six other OECD countries (Japan, West-Germany, France, Great Britain,Italy and Belgium). The ratio is exceptionally high for Sweden in industries dominated by largecompanies. Unfortunately, there are only figures for 1985 and 1987. Statens Industriverk [1990, p.82].

48 Carlsson [1988, pp. 22-23], Statens industriverk [1988, p. 63] and Giertz [1991, pp. 257-262].

49 Dahmén, 1950 [1950, pp. 360-367 and 381] and Giertz [1991, p. 150]. European countries ingeneral had been slow to catch-up new U.S technologies in the interwar period (Fagerberg [1994, p.1158]).

50 The strategy seems to contradict the Swedish innovation model where large companies´ entries innew industries are noticed. However, new industries were mainly penetrated by the most home-marketorientated companies, Saab(-Scania), Volvo and ASEA, in the late Golden Age. In fact, ASEAcombined a “diversification” and core strategy by both expanding in new industries (inter aliaindustrial robots) and putting a stronger emphasize on established products. ASEAs strategy impliedthat the company closed or sold units operating on stagnating markets or with weak technologicallinks to the rest of the company.

33

to the core business. In the course of time, some of the companies had become generaldealers rather than specialists.51

It is true that the core strategy of Swedish companies increased their dependency ofsingle products. Besides, several core products experienced a relatively low marketgrowth in the Golden Age. But the companies expected that any tendency of marketsaturation would be counter-balanced by a stronger market position. The hopes werenot ungrounded. The Swedish companies had procured a large capability to compete inparticular (even standardized) segments of the world market.

Tendencies towards market saturation for core products might have contributed, asharder international competition, to R & D and rationalization efforts by large Swedishcompanies in the 1960s and early 1970s. Such tendencies also induced the companiesto adjust to fast-growing regions abroad and to search for new uses of their coreproducts. It seems that the large Swedish companies in engineering were successful inthese respects.52

51 Alfa Laval sold its steel pressing company Olofström AB (to Volvo) in 1969. Furthermore, AGAabandoned in the early 1970s its postwar strategy to diversify outside its “core”. The company hadbecome a producer of medical equipment, home-electronics, radiators and batteries in the 1960s.ASEA sold its household product company, Electro-Helios, to Electrolux already in 1962.Furthermore, ASEAs production of welding equipment and elevators was laid outside the company inthe 1960s although ASEA maintained its control by remaining as a predominant owner.The “core strategy” of the large companies in the late Golden Age implied in some cases that anearlier specialization pattern was confirmed. SKF experienced tendencies towards market saturationand harder international competition (primarily through a Japanese challenge) in the early 1970s. Thecompany´s strategic decision was to remain in ball-bearings, not to broaden its product programme.Atlas Copco had already specialized in pneumatic products in 1948 when the company abandoned its(unprofitable) production of diesel engines. Technological progress by competitors in the 1950s,eliminating the comparative advantages of “the Swedish method”, did not lead to a decision by AtlasCopco in the 1960s to market products alien to the core business.Electrolux seems to contradict the thesis of product specialization. The company expanded in marketsfor white-goods and cleaning and garden equipment in the 1960s and early 1970s. In fact, stoves werenot included in the company´s product programme before the requirement of Electro-Helios. However,Electrolux´ ambition was to enlarge its household assortment to products with strong commercial andtechnological links to its “core” and to become a Swedish (and Scandinavien) monopolist for whitegoods. Electrolux´ core strategy is confirmed by the fact that it closed down a “peripheral” productionof warm-water radiators (inherited from Electro-Helios) by selling the business division to AGA in1971.All large companies in Swedish engineering did not choose a specialization strategy in the 1960s.Swedish Match´s ambition to diversify (mainly through take-overs) in the Golden Age was raised inthe late 1960s. Besides, the merger between Saab and Scania-Vabis was inter alia an attempt toreduce the companies reliance on private passenger cars and trucks/buses respectively. However, adecision had earlier been made by Scania-Vabis to abandon a diversification into light trucks, tractorsand cars.

52 Estimations of changes in countries´ share of world trade do not unveil a strong regional flexibilityof Swedish engineering in the late Golden Age (Mjøset et al [1997]). But the estimations are tooaggregate to discern the flexibility of leading companies. Firm-specific information shows thatcompanies such as Atlas Copco and AGA succeeded to find new industrial applications of their coreproducts. Furthermore, ASEA reduced, inter alia by product differentiation, its reliance ongovernment authorities in energy in the 1960s by significant sales to industries, households and publicauthorities in transport.

34

Limitations of the Swedish market led, together with trade liberalizations (and otherreductions in trade barriers), to a strong increase in sale efforts abroad by largeSwedish companies. The foreign share of their total sales was raised considerably inthe 1960s, see table 10. The increase was remarkable for two companies with a strongreliance on domestic markets in the early Golden Age, Volvo and ASEA.

Volvo and ASEA first became outward in the 1960s and early 1970s if we classify acompany as outward when foreign sales exceed domestic sales. Volvo, and also Saab,could benefit from the fact that world demand growth for cars became higher than formost other products in the late Golden Age (Mjøset et al [1997]). Swedish producersof private passenger cars succeeded to make significant inroads into the U.S marketalready in the second half of the 1950s. Volvo became Sweden´s largest exportcompany in the early 1960s.53

The success of the large Swedish companies in the Golden Age has been related totheir early “outwardness”. A long experience of competition on foreign markets and ofestablishing subsidiaries abroad gave a real competitive advantage of Swedishcompanies.Virtuous circles did also arise in the late Golden Age as the earlier successof Swedish companies facilitated the financing of take-overs and of investments in R &D and marketing.

It is true that Saab and Scania-Vabis were successful in the Golden Age without anylong experience of export. But they recruited managers from the large Swedishtransnational companies (Giertz [1991, p. 297-298 and 325-326]). In addition,industrial capitalists such as the Wallenberg family acted as an intermediater ofexperiences from early outward companies to companies which were latecomers on theworld market.

But it is wrong to reduce the success of Swedish companies in the Golden Age to onevariable only, “early outwardness”. References must be made also to high managerialskill (unrelated to foreign experiences) especially in the case of Electrolux from the mid1960s. Furthermore, the major industrial owner, the Wallenberg family, permitted longinvestment horizons in the leading Swedish companies. The family´s position as majorowners also facilitated the merger between Saab and Scania-Vabis. The validity of anownership theory of the Swedish success is, however, restricted by the fact thatcompanies outside the Wallenberg sphere, often with more passive owners, such asAGA and Volvo were fortunate (and foreign orientated) as well.

Political factors contributed to the success of large Swedish companies in the GoldenAge. The Swedish innovation model offers one political explanation. Productdevelopments induced by qualified public demand, particularly in Ericsson, ASEA,Volvo, Scania-Vabis and Saab, opened the gates to international markets. Theneutrality of Sweden was another factor behind the success of Swedish companies. Itmade it easier for Swedish companies to get foreign orders from and establish foreign

53 In the late 1950s, private passenger cars became Volvo´s main products (in value terms) even iftrucks and buses are put together.

35

subsidiaries in underdeveloped countries. Thus, the importance of “luck” for theSwedish success shall not be completely ignored.

Firm size is another explanation of why Swedish companies in enginering werefortunate in the Golden Age. Small size was not always a disadvantage of Swedishcompanies as it often made them more flexible. For instance, contacts betweenmarketing, R & D and production departments were facilitated. Such contacts made iteasier for Swedish companies to adjust research and production to new demandpatterns.

5.2 Threats, Opportunities and Reactions II - The Raw Material Industries

Sweden´s specialization on raw material industries resulted in high export growthimmediately after World War II. The industries gained by excess demand and highproduct prices especially during the Corean War.

However, the Swedish saw mills and steel producers were soon exposed to hardinternational competition. The Swedish steel companies were hit by foreigncompetitors´ use of new technologies already in the late 1950s. They had, as theSwedish saw mills, a vulnerable vintage structure as investments by foreigncompetitors which had suffered from the war embodied the latest technology. Thenumber of plants in Swedish wood product and steel industries decreased sharplyalready in the late 1950s.

The competitive pressure on Swedish iron and steel industries was intensified in the1960s. In particular, Japan and, later, newly industrialized countries, began topenetrate Western markets for iron and steel products.

The Swedish iron and steel industries and saw mills were accustomed to hard foreigncompetition. Already in the 1890s, Swedish iron and steel industries had pursuedquality up-gradings to meet competition from larger European countries whichbenefited from cheaper inputs and economies of scale. Similar up-gradings tookplace, although with a delay, in saw mills, inter alia through integration with pulpproduction.54 In the 1920s, tough international competition led to both productdifferentiation and the introduction of new technologies in saw mills and in iron andsteel industries.55

In contrast, Swedish pulp and paper industries had not faced any strong internationalcompetition before World War II. In fact, they were first hit by a strong competitivepressure in the mid 1960s. Now, North American competitors, helped by reductionsin sea-transportation costs, invaded European markets. Above all, North Americanproducers sold off surpluses when demand was weak on their domestic markets.

54 Dahmén [1950, pp. 169-171 and 304-310], Jörberg [1961, pp. 70-71 and 83-84] and Glete [1987,pp. 198-204].

55 Dahmén [1950, pp. 127 and 157-161].

36

Furthermore, Finland - the main European competitor to Sweden in pulp and paperindustries - devaluated in 1967.

Productivity growth was raised considerably in Swedish raw material industries in the1960s.56 High demand growth in the first part of the 1960s offered greatopportunities to use scale advantages. It also led to high investments embodying newtechnologies (both in older and newer plants).

The ability of the raw material industries to show high productivity growth wasreduced during the second part of the 1960s by successively higher productivitylevels. Moreover, the possibilities to exploit scale advantages were, ceteris paribus,impaired by a lower demand growth. But productivity growth was sustained in basicindustries mainly because of rationalizations (including mergers), plant specializationand eliminations of inefficient vintages.

It seems that transformation pressure rather than new opportunities (commercial,technical or financial) was the driving force behind the large productivity increases inSwedish raw material industries in the latter part of the 1960s. Pressure was raisedprimarily through tougher international competition and lower demand growth.57 Aharder international competition was partly triggered off by a an internationalrecession in the mid 1960s (cf. pulp industry above). There was also a stronger wagepressure on some raw material industries (mainly saw mills) through wage leadershipby large export companies, solidaristic wage policy and a growing labour scarcity.

A merger (and take-over) boom occurred in Swedish manufacturing particularly inraw material industries during the second half of the 1960s. Mergers contributedstrongly to productivity growth in manufacturing in the 1960s.58 In the raw materialindustries, mergers were primarily induced by tight international competition andrecessions (Rydén [1971, pp. 198-206]). The potential for mergers was large sincethey had been postponed in the 1950s due to high profitability.

The room for “enforced” productivity increases was huge in general in Swedish basicindustries in the mid 1960s particularly in pulp and paper industries where thecompetitive pressure had been weak for so long (Wohlin [1970, p. 109]). Paper andpulp producers could handle harder international competition rather efficiently in spite

56 It seems that the increase in labour productivity growth was explained more by a higher growth incapital intensity than by an increase in total factor productivity growth (cf. Wohlin, [1970, p. 108]).However, mechanisation is often associated with technological progress.

57 There was no significant positive relationship between international competition and the use ofproduction slacks (X-inefficiences) in 1968 according to a cross-sectional analysis of 26 Swedishmanufacturing industries (Carlsson [1972]). However, the argument about rationalizations aboveconcerns changes during the Golden Age and raw-material industries only. Besides, the measures ofinternational competition at the study were rather crude.

58 Estimations have shown, that 25 per cent of productivity increases in Swedish manufacturing 1966-1969 is explained by mergers - the corresponding figures for the 1951-1969 period are 15 per cent.The raw material industries´ share of total mergers in Swedish manufacturing 1966-1969 was 35 percent if printing and publishing branches are included, see Rydén [1971, pp. 55 and 148].

37

of weak experiences of such pressure. Productivity growth was higher in pulp andpaper industries than in other basic industries in the second half of the 1960s.

The conclusion that Swedish pulp and paper industries were able to meet thechallenges of the 1960s is not obvious. First, their high productivity growth was to alarge extent explaned by “capitulate” plant closures.59 Second, the potential forproductivity increases was large. Third, the pulp and paper industries were the onlyindustries in Sweden, beside ship building, which experienced large losses of marketshares in the second part of the 1960s.

But the erosion of market shares in pulp and paper industries was larger for Finlandthan for Sweden. Swedish pulp and paper producers did also regain large sharesbefore the Golden Age was over.60

Finally, plant closures and losses of market shares for raw material industries wereclear expressions of structural change in Swedish industry. There were no large lossesin aggregate market shares for Sweden in the 1960s and early 1970s in comparison tosubsequent periods. The total losses of world market shares in the second part of the1960s were of equal size for Norway and Finland and much larger for Denmark (andIceland), Erixon [1988, pp. 13] and Mjøset et al [1997].

Another indication of structural change in Sweden during the late Golden Age wasthe switch in export from staple raw-materials to finished goods. The group ofunfinished goods, according to the SITC classification in trade statistics (group 1-4),includes agricultural products, wood products from saw mills, pulp and ore. Thefinished goods (SITC group 5-9) raised their share of Swedish export with 9percentage points 1963-1973 in spite of a “raw material boom” in 1973. The increasewas larger than in most other OECD countries (Horwitz [1979, p. 43]).

59 According to Lars Wohlin, the increase in productivity growth in Swedish pulp industry during thesecond half of the 1960s is fully explained by the fact that more plants with low productivity wereclosed (Wohlin, [1970, p. 109]). Official statistics show, however, that labour productivity growth inpulp and paper industries was least as high in the first as in the second part of the 1960s (Erixon[1987, p. 223]).

60 Mjøset et al [1997] and Erixon [1988, pp. 13-17)]. Iron and steel industries must be added to thegroup of Swedish industries with large losses of market shares if only the 1969-1973 period isconsidered.

38

Chapter 6 - The Coherence of the Swedish Model

The leading industrial sector succeeded to establish high growth in Sweden in theGolden Age. The sector showed high export and productivity growth and was ablealso to both generate strong demand multiplier effects and to spread new knowledge.New ideas were primarily produced or transferred by the large companies inengineering. For instance, the ideas of scientific management, applied by companies ofthe Swedish model already in the interwar period, were rapidly spread in Swedishmanufacturing after World War II (Giertz [1991, pp. 158-159]).61

The success of the large companies and dominant export branches was a necessarycondition for the coherence of the Swedish model. But the postwar model alsoassumes that profitability in the leading industrial sector was kept up by economicpolicy. In fact, Saab, Volvo and Electrolux would hardly had grown to largecompanies in the Golden Age without a prosperous combination of public orders,regulations, welfare reforms, tax advantages and expansive economic policies.

A statement that capital accumulation in the leading industrial sector was supported byeconomic policies in the Golden Age must, however, be qualified. First, there wereother domestic driving forces at the macroeconomic level. Independent of economic(and welfare) policies, there was a strong demand for private passenger cars, housing,household appliances and home electronics (wireless and television sets andgramophones). Household demand was particularly strong immediately after the warbut it was stabilised on a high level by high birth-rates, rising real wages and the break-through of new social values, “the American Way of Life”.

Second, the contributions from economic policy to capital accumulation in the GoldenAge did not mean that policy makers in Sweden were put under a permanent pressurefrom leading companies and share holders to offer them privileges. A theory thateconomic policy measures were enforced by the leading industrial sector is, forinstance, refutet by the fact that large engineering companies were reluctant to aSwedish devaluation in 1949.62

It is true that formal meetings were held 1960-1964 between the government andrepresentatives from big business, a co-operative arrangement called “Harpsund

61 However, corporatist arrangements in the labour market contributed independently to the extensiveuse of scientific management in Swedish manufacturing in the Golden Age. A judgement by theLabour Court in 1944 made piece-work studies compulsory in Swedish enginering (Giertz [1991, pp.160-161]).

62 Söderpalm [1976, p. 150]. The opinion that a devaluation would lead to “export without effort” wasalso shared by the Wallenberg family according to unofficial statements by the family´s economicadvisor in the Golden Age, Erik Dahmén.

39

Democracy” in Sweden. But the real impact of the Harpsund meetings should not beexaggerated.63 More important, there was a strong conviction in Sweden amongleading Social Democrats and trade unionists in the Golden Age that growth wasstimulated by export, large companies and industrial concentration.64

Besides, the biased character of the profit tax system were not always known bypoliticians or governmental experts and the priorities of full employment andredistribution were seldom related to industrial dynamics. Keynesianism was a shortrun strategy to obtain a full utilization of human and physical resources, not a tool todevelop a Swedish mass-consumption industry in the interest of labour and largecompanies. Swedish economic policies in the Golden Age cannot be “explained”entirely by regards to dominant business interests.

But I will not abandon a theory that the leading industrial companies had an impact oneconomic policy in the Golden Age. First, at some occasions, the leading companiespersuaded the government to offer them favours (Söderpalm [1976, p. 150]). Second,the lack of a strong pressure from industry on policy makers is, in fact, a confirmationof a theory that economic policy was determined by the interests and attributes of theleading industrial sector. The sector was differentiated enough to prevent that specificexport interests became decisive for economic policy and that serious economicimbalances appeared in Sweden raising political claims of ad hoc measures to favourthe sector.

For instance, Sweden´s more differentiated industry explains the absence of aninflationary Finnish (or Icelandic) devaluation cycle in the Golden Age. Swedisheconomic policy could not be adjusted completely to the claims of the raw materialindustry. Moreover, Sweden´s diversified export sector prevented large current-account deficits and losses of market shares during the second half of the 1960s. InFinland, the need of a devaluation was stronger as raw materials had a larger weight ofexport.

6.1 The Embryo of the Swedish Model

The Swedish model did not suddenly appear after World War II. Its birth can be dated,with some reservations, to the 1930s. The Social Democrats formed stablegovernments with a parliamentary support from the Agrarian Party from 1932 and theLO was ultimately considered as an established social partner through the Saltsjöbadenagreements. In addition, economic policies in the 1930s were benevolent for industrieshere associated with the postwar Swedish model.

63 The leading industries had become more antagonistic in the late 1940s as there were someindications of an active growth policy by the Social Democrates and propositions of higher profit andwealth taxes. However, the militancy of the large outward companies was soon replaced by a more co-operative attitude primarily reflecting a Social Democratic return to a passive growth policy.

64 This is illustrated by the postwar economic programme of the Swedish labour movement (1944) andthe subsequent discussion among leading Social Democrats about industrial and employment policies.

40

Independent of Social Democratic policies, the SEK was depreciated 1931-1933 (afterthe abandon of the Gold Standard). The depreciations were favourable for Swedishraw material industries.65 They did also benefit the new consumption industries. Anundervalued SEK resulted in a strong comparative advantage of Swedish producers ofconsumer goods in home markets. In the 1930s, total Swedish import increased with17 per cent (value) while Swedish import of consumer goods decreased with 15 percent (Lundberg [1945, pp. 22-23]).

The depreciations did also stimulate domestic demand for consumer goods by makingthe Swedish economy more liquid. The pressure “upwards” on the rate of interest toobtain balance of payment was easened. Lower interest rates promoted demand fordurable consumption-goods both directly and indirectly by higher investment inconstruction.

The consumption industries were favoured also by a counter-cyclical, generallyexpansive, fiscal policy from the early 1930s.66 The more expansive effect of fiscalpolicies in the 1930s compared to the 1920s reflected higher public sector growthrather than larger budget deficits (Bergström [1969, p. 49]). Demand for durableconsumption goods was stimulated and stabilized by pension reforms, householdtransfers (including unemployment insurances) and housing subsidies.

The importance of fiscal policies for aggegate demand in Sweden in the 1930s shouldnot be exaggerated. Depreciations of the SEK and expansive monetary policies hadprobably stronger effects on GDP and employment. Besides, the depreciations maderoom for expansive fiscal and monetary policies.67 In addition, real wage increaseswere more important for consumer demand than fiscal policies. High real wage growthin the 1930s did not primarily reflect strong labour (see the Swedish model) but thediffusion of new technologies.68

Depreciations and expansive economic policies led to high domestic demand forSwedish consumer-good companies in the 1930s. Morover, foreign penetrations ofSwedish markets (and a more export-orientated strategy of Swedish companies) wereblockaded by protectionism. Fierce foreign competition on domestic markets mighthave destroyed the fragile Swedish consumer-good industries in their childhood.Together with protectionism, Keynesianism (and depreciations) was probably the rightthing in the right time for companies in the second innovation wave.

65 Depreciations did also occur in other Western countries but not as early as in Sweden (or GreatBritain).

66 Fiscal policies were restrictive in the second half of the 1930s to avoid overheating.

67 Improvements in the current account, inter alia reflecting an undervalued SEK, did also lead toincreases in liquidity and reductions in the rate of interest when the SEK was finally tied to the poundin 1933 (Bergström [1969, pp. 31 and 45]).

68 Membership in the LO increased rapidly in the interwar period (D´Agostini [1987, pp. 16-18]). Butthe rate of unemployment, a valid determinant of labour strength, was above 9 per cent even in thelate 1930s.

41

Expansive economic policies in the 1930s did also stimulate companies in the firstinnovation wave. This was particularly true for producers of infra-structural andconstruction equipment such as ASEA and Atlas Copco. However, for Atlas Copco,the home-market orientation of the 1930s was a temporary departure from afundamentally outward strategy, not a natural phase in the company´s product cycle.

Profitability and profit shares (of value added) in Swedish manufacturing increased inthe 1930s (Jungenfelt [1966, pp. 35-36]) and Wibe [1976, p. 27]). The increasereflected a favourable combination of high foreign demand, selective protectionism69,expansive economic policies, depreciations and weak labour. Furthermore, industrialprofits were raised rather than squeezed by the diffusion of new technologies. Inaccordance with the Swedish innovation model, new knowledges were primarilyspread from one industry to another (Dahmén [1950, 378-379]).

Higher profits largely explain why internal saving became the most important source ofindustrial investment in the 1930s (Dahmén [1950, pp. 380 and 396]). The importanceof plough-back profits was raised by a new corporate tax system in 1938 where freedepreciation allowances for machines and equipment were permitted. The Swedishgenerosity of tax allowances was striking even in an international perspective. The newcorporate tax system was particularly benevolent for large and capital-intensivecompanies. It virtually weakened the outward large companies´ hostility to SocialDemocratic governments.70

In the 1930s, it is too early to talk about a Swedish model as defined in chapter 2.Labour was still too weak and consumer-good companies were not yet able to work asgrowth engines. But stabilization and tax policy routines of the early postwar periodcould already be distinguished in the decade.

6.2 Economic Policies in the Golden Age

Economic policies in the Golden Age had the same parliamentary basis as in the 1930s.Social Democrats formed governments during the whole period. The Agrarian Party(later the Centre Party) joined the Social Democrats in a coalition 1951-1957.

6.2.1 The Early Golden Age

69 Trade restrictions favoured Swedish home market industries but they did not prevent Swedishexport of raw materials.

70 The managers of ASEA, Electrolux, LM Ericsson, Separator (Alfa-Laval) and SKF were membersof a political pressure group 1933-1953. (AGA joined the group in 1941.) The group first opposed theKeynesian policy of the Social Democrates but also the “neutral” employer strategy manifested by theSaltsjöbaden agreement, Söderpalm [1976].

42

After World War II, the raw material industries were favoured once again by economicpolicies. The devaluation of the SEK in 1949 engendered a comparative advantage ofSwedish basic industries (and mining).71 In particular, it delayed the invasion ofWestern European markets by North-American pulp producers. Furthermore, theabsence of a revaluation of the SEK during the Korean War created a profit boom inSwedish raw material industries.

The raw material producers benefited also, at least up to the mid 1950s, by generallylow rates of interest and, during the whole Golden Age, by the fact that falling energyprices were not neutralised by tax increases (cf. Schön [1990, pp. 30-33 and 75-76]).Finally, capital-intensive industries such as the raw material ones were not seriously hitby “full” employment policies (se below).

Swedish engineering gained from the 1949 devaluation as well. In fact, as in the 1930s,Swedish producers of consumption goods were favoured by a combination of anundervalued SEK and import restrictions.

Import restrictions were implemented in Sweden in 1947 to reduce current-accountdeficits caused by expansionary monetary and fiscal policies (and a revaluation of theSEK the year before). The restrictions, which were not removed until 1954, were ofoutmost importance for the Swedish car industry. They led, together with the 1949devaluation, to a strong reduction in import growth of both cars and of componentsused by foreign car producers with assemblies in Sweden. For instance, car import (invalue) decreased by 7 percent 1950-1952 when total import increased by 47 per cent -table 14. As a consequence, Volvo and Scania-Vabis got a strong position on Swedishmarkets for trucks which they sustained during the whole 1950s. Import restrictionsdid also favour Swedish producers of private passengers cars since domestic demandwas decisive in their first expansive phase until the late 1950s. They contributedstrongly to the fact that Volvo replaced Volkswagen as the most sold car in Sweden in1956.72

Import restrictions did also benefit Swedish suppliers in car industry. A large numberof new supplying firms, entirely depending on orders from, Volvo, Saab and Scania-Vabis, was established in the late 1940s and the early 1950s (cf. Giertz [1991, pp. 271-273]).

It is not obvious that Swedish producers of consumption goods did benefit, as the rawmaterial producers, from the 1949 devaluation. It led to a decline in real wages ceterisparibus hampering demand for consumer durables. Besides, the export share of totalsales was lower and the import share of total inputs higher for engineering companiesthan for raw material producers. But the undervalued SEK lead to significant (netexport) demand multipliers and to competitive advantages of Swedish companies in

71 The SEK was evaluated by 30 per cent in relation to the U.S dollar, following a devaluation of thepound sterling. (A 17 per cent revaluation of the SEK had been made in 1946.) As other countriesfollowed the British example, the SEK was devaluated in practice with 13-15 per cent.

72 Similar import restrictions in other countries postponed the export of Swedish cars, particularly toother Nordic countries.

43

engineering, particularly in relation to American producers. The devaluation did alsolessen the external restrictions on Swedish monetary and fiscal policies in the 1950s, aclear advantage especially for the consumer-good industries.

As already mentioned, the interest rates were permantly low, at least until the mid1950s, thus, monetary policies were passive from a stabilization viewpoint.73 The realrate of interest shrank steadily in the 1950s until the economic upswing at the end ofthe decade. Attempts were really made until the recession in the late 1950s to restrictcredits by agreements between the Central Bank and commercial banks and the use ofnew rationing instruments (see below). However, monetary policies did only becomerestrictive during some years in the mid 1950s (Bergström [1969, pp. 117-123] andJonung [1993, pp. 295-300]). Besides, the leading companies/industries were probablyless hit by capital rationing (cf. 7.4.1).

From the end of the war to the late 1950s, Swedish fiscal policy may be dubbedKeynesian. It was clearly counter-cyclical and expansive in general (Matthiessen,[1971, p. 176]). As a consequence of fiscal and welfare policies, demand for consumerdurables was stabilized on a high level. Besides, a rise in investments´ share of publicexpenditures until the mid 1950s was beneficial for companies such as ASEA, Scania-Vabis, Volvo and Atlas Copco.74

Fiscal attempts were made until the late 1950s to restrict private investment by fees,investment taxes, higher statutory corporate taxes, stricter rules of inventory valuationand a phasing out of the use of free depreciation allowances for machinery andequipment (Bergström [1969, pp. 66-67] and Södersten & Lindberg [1983, p. 41]).However, the total tax burden on companies was stable until 1958.75 In addition, taxrestraints were counter-balanced by huge depositions in investment funds from the mid1950s which were particularly beneficial for large and capital-intensive companies (seenext section). Furthermore, the measures to curb private investments were offset byexpansive fiscal (and welfare) policies leading to demand-induced investments,particularly in consumer industries. Besides, labour-intensive producers of consumerdurables were less hit by fiscal measures to restrict investments.

Profitability (both before and after taxes) and profit shares in Swedish manufacturingwere generally high in the late 1940s and the 1950s at least in comparison tosubsequent decades (Erixon [1987, p. 43] and [1991, p. 282]). There was only one“branch of crisis” - textile and garment industry. Profits were high in spite of a growinglabour shortage. They reflected favourable demand conditions and a vigour of leadingindustries independent of “policies” but also an undervalued SEK, expansive fiscal and 73 Jonung [1993, pp. 295-298]. A Keynesian label on Swedish monetary policies in the early GoldenAge is justified by Keynes´ recommendation in the General Theory that the rate of interest should bekept low over the business cycle.

74 Figures on total public investments and total public expenditures 1950-1973 have been broughtfrom Statistics Sweden [1990, table 11.7].

75 See figures on effective tax rates (taxes actually paid as a ratio of operating surplus less capitalconsumption) in manufacturing and on companies´ share of total taxes in Sweden, Erixon [1987, pp.42-43)] and Bergström [1969, p. 116].

44

monetary measures and trade regulations. For instance, a fragile Swedish car industrywas first protected by import restrictions. The similarities between the 1930s and theearly Golden Age are striking.

It seems that economic policy (broadly defined) in the early Golden Age wasparticularly advantageous for the industries of the Swedish model. For instance, theywere supported by profit taxes, at least from the mid 1950s. As economic policiescontributed to high profits in the leading industrial sector, they were also conducive tothe sector´s further expansion. The leading companies´ ability to build up internal riskcapital and attract equity capital was raised. The discriminatory effects of these policiesare discussed in section 7.4.1.

6.2.2 The Late Golden Age

Swedish fiscal policies were still predominantly expansive (mainly as a reflection ofpublic sector growth) from the end of the 1950s. In addition, fiscal policies hadcounter-cyclical effects, at least until the mid 1960s (Matthiesen [1971, p. 176] andLindbeck [1975, p. 104]). This was also true for monetary policies which now becameinstruments of Swedish stabilization policies (Jonung [1993, pp. 298-303 and 346-348]). Beside discount policy, a broad range of credit regulating devices were used tooffset business cycle fluctuations such as interest rate penalties, rules for commercialbanks´ liquidity and cash accounts and ceilings for commercial lending. Largervariations in the real rate of interest from the mid 1950s reflected (although notwholly) a more counter-cyclical monetary policy.76

At first glance it seems, that Swedish economic policies in the late Golden Age wereless advantageous for the leading business sector. Pre-tax profitability and profit sharesfell in most branches of manufacturing in the 1960s and early 1970s. In fact, Swedenexperienced, together with Great Britain and the Netherlands, a stronger decline inprofit shares in manufacturing than other OECD countries (Erixon [1987, p. 53] andErixon, [1994, p. 38]). Moreover, the Swedish decline in pre-tax profit shares wasworser for larger than for smaller firms (Södersten [1971, p. 323] and SverigesVerkstadsförening ([1980, pp. 31-34]).

The tendency of a profitability decline in Swedish manufacturing could have beenabolished, at least temporary, by a devaluation exactly as in Denmark, Finland andGreat Britain (1967). A devaluation would also have reduced the current-accountdeficits emerging during the peaks of the mid 1960s and the late 1960s. But deficits inthe current account, and also in the capital account, were met by restrictive monetaryand fiscal policies, not by a devaluation. Credit restraints were also used to preventinflation in the economic upswing 1969-1970. As a result, monetary policies becamemore restrictive than ever before in the postwar period with a possible exception of themid 1950s. 76 The enlarged use of monetary instruments in stabilization policies in the 1950s was no uniqueSwedish feature. The same development can be distinguished, for instance, in other Nordic countries.

45

Furthermore, counter-cyclical fiscal measures were applied too late in the peaks 1964-1965 and 1969-1970. As a consequence, fiscal restraints hit the Swedish economy atfull strength in the following recessions. Moreover, restrictive fiscal (and monetary)policies were needed during the first part of the recessions to eliminate current-accountdeficits.

Thus, partly through bad timing, partly through a reluctance to devaluations, monetaryand fiscal policies became restrictive during or immediately before the recessions in themid 1960s (close to neutral in the fiscal case) and the early 1970s. In general,stabilization policies became more restrictive from the mid 1960s. Besides, as far asfiscal policies are concerned, investments´ share of public expenditures decreasedsharply in the mid 1960s. This was inauspicious for the large Swedish producers ofinvestment goods.

But the conditions for accumulation in the leading industrial sector in the late GoldenAge were not bad after all. Among industries in Swedish manufacturing, saw mills andpulp industries only were hit by a sharp decline in pre-tax profitability until 1973. (Theiron and steel industries must be added to the group if profit margins are accountedfor.) Besides, the profitability decline in large wood companies seems to have beenmodest (Rydeman [1980, p. 132]).

Producers of transport equipment except shipbuilding - a branch dominated by Saab(-Scania) and Volvo - did not experience any decline in pre-tax profitability (or profitshares) at all. Swedish companies in engineering, particularly in transport equipmentindustry, were helped by a German revaluation in 1969. Moreover, they still benefitedfrom the fact that investments´ share of public expenditures had increased from the mid1950s to the mid 1960s. Improvements in Swedish high-way systems led to a strongincrease in domestic demand for heavy trucks (Giertz [1991, pp. 283-284]).

In addition, the profit tax system effectively mitigated the fall in profitability of theleading industrial sector. The decline in profitability in Swedish manufacturing as awhole 1953-1972 is halved if profit taxes are accounted for (Erixon [1987, pp. 38-39]).77 The reduction in effective profit taxes was primarily due to depreciationallowances (at high investments) and the investment funds. 78

77 The effective tax rate in Swedish manufacturing was significantly higher in the 1950s than in the1960s and the early 1970s. However, the sharpest decline occurred in the second half of the 1950sErixon ([1987, pp. 42-43]).

78 A reduction in statutory corporate tax rates in 1960 was of less importance (Södersten [1971, pp.226-332]). International comparisons of the late Golden Age do not unambiguously show that profittaxes actually paid were lower in Sweden than in other OECD countries. In 1960, corporate taxes as aper centage of GDP was the same in Sweden (2, 4 per cent) as in the present EU countries on averagebut much lower than in the OECD on average (3, 8 per cent), see Hoeller et al [1996, table 9].Furthermore, it seems as if the Swedish case of a milder profitability decline after taxes hascorrespondences in other countries (Erixon [1987, p. 52]). A conclusion that Swedish corporate taxeswere not lower than in other OECD countries in the late Golden Age is confirmed by a comparativestudy of effective marginal taxes in 1980 in Sweden, Great Britain, West-Germany and the U.S(Södersten & Lindberg [1983, pp. 30-32]).

46

The use of investment funds since the mid 1950s to reduce business cycle fluctuationswas a typical Swedish innovation. Swedish firms were already through the profit taxreform in 1938 allowed to make deposit deductions from their profits in particularfunds for investments. Deductions up to 40 per cent were permitted in the GoldenAge. From the mid 1950s, the main condition for allowances was that a certain shareof the deduction (almost half of the deduction in the Golden Age) must be placed ininterest-free accounts in the Central Bank. These reserves could be regained dependingon a permission by the Central Bank following an assessment in the light of the goalsof stabilization policy (Södersten & Lindberg [1983, pp. 53-56]).

The investment funds had a strong positive effect on the consolidation of largecompanies and capital-intensive industries (Södersten [1971, pp. 324-325 and 329]).Tax allowances for large companies in engineering raised financial funds not only forinternal expansion but also for take-overs.79

The leading industrial sector was not hit by restrictive monetary policies in the late1960s and early 1970s. Increases in the rate of interest did not lead to strong increasesin the sector´s capital costs because of tax deductions. In addition, export companieswere excepted from monetary restraints inter alia through permissions to borrowabroad. The old outward companies had the largest opportunities to take foreign loanson favourable terms. Moreover, industrial companies circumvented monetaryrestrictions by borrowing on “grey” credit markets in Sweden and by extensive selffinancing.

To sum up, because of accumulated savings, tax allowances and favourable foreignloans, investments by the large outward companies were not seriously restrained bymonetary policies in the late 1960s and early 1970s. In Swedish manufacturing,medium-sized firms only were gravely hit by deflationary monetary policies.80

The large companies in engineering benefited in the late Golden Age not only from taxallowances and macroeconomic policies but also from active labour market policies.Selective job creation programmes and measures to improve matching and affectlabour supply were consistently implemented during the 1957-1958 recession andgained substantial importance in the mid 1960s.81 At the end of the Golden Age,

79 In 1960 and in 1970, the statutory corporate tax rate in Sweden were 49 and 53 percent and theeffective corporate tax rate at investment funds 37 and 41 percent respectively (Södersten & Lindberg[1983, pp. 53-56]).

80 Matthiessen [1971, pp. 211-213], Kragh [1973, pp.156-158], Bergström [1981, p. 3], Södersten &Lindberg [1983, pp. 30-32] and Wihlborg [1993, pp. 266-267 and 254].

81 Two trade union (LO) economists, Gösta Rehn and Rudolf Meidner criticized the stop-goKeynesian policy of the 1940s and the early 1950s. In a report to the LO congress in 1951, theysuggested a programme for obtaining full employment without inflation. Contractive economic policy(primarily fiscal policies) should control inflation by squeezing profit margins. Full employment wasto be obtained by labour market programmes (training, subsidies for occupational and geographicalmobility, regional policies and labour market information). They assumed that such measures wereless inflationary than expansive fiscal and monetary policies.

47

expenditures on active labour market policies were higher in Sweden than in otherOECD countries (Erixon [1984, pp. 10-11]) and SOU 1993:43 [1993, p. 125].Supply-orientated measures to stimulate occupational and regional mobility expandedmost from the late 1950s to the early 1970s (Johannesson [1981]). The supplymeasures benefited large engineering companies which faced a growing labourshortage and were mainly located in the southern regions of Sweden whichexperienced a large net immigration in the 1960s (cf. Nilsson [1995, pp. 19-22]).

Standardized producers of consumer goods benefited from welfare reforms in the lateGolden Age. Real investment in housing increased steadily from the early 1950s to themid 1960s to culminate during the time of the Million Programme. The incrementalpension funds (the ATP) were the most important financiers of the programme in termsof bond holdings (Englund [1993, pp. 161 and 190-191]).

Swedish consumer-good industries benefited also from redistribution policies.Narrowed wage and income differentials after taxes in the 1960s speeded up the use ofdurable consumption-goods. Redistribution policy partly explains why privatepassenger cars from Volvo and Saab became common properties of Swedish familiesin the late Golden Age.

Trade union policies from the late 1950s were clearly in line with the interests of theleading industrial sector. Solidaristic wage policy (including policies to level out wagedifferentials in general from the mid 1960s) did not seriously hit the leading industrieswith the exception of small (labour intensive) saw mills. The strength of the LO in the1960s did not challenge the wage leading role of the large export companies.

Solidaristic wage policy stimulated demand for products of large engineeringcompanies. It did also stimulate, as employment policies, their product development.The incitements to use capital-intensive techniques in Swedish manufacturing wereraised due to higher costs of unqualified work. For instance, the use of industrialrobots in Swedish engineering increased in the 1960s. As a consequence, strongincitements arose in manufacturing to supply other industries with labour savingequipment (Schön [1990, p. 82]). ASEA became the largest domestic supplier and alsoa world leading producer of industrial robots in the 1970s.

My view of Swedish wage formation excludes the possibility that the leading sectorwas restricted by wage policies in the Golden Age but also, that they gained by “wagemoderation” as a result of solidaristic wage policy82 or commitments by the LO topreserve Swedish competitiveness. The large export companies were wage-leaders. Itis true that solidaristic wage policy hit the leading companies indirectly by leading tohigher input prices. But positive indirect effects of solidaristic wage policy must beconsidered as well. The policy speeded up the demand for consumer durables in lowerincome brackets and for labour-saving techniques. Such demands were exploited bylarge companies in engineering.83

82 According to the theory of solidaristic wage policy, profitable firms are favoured since they are ableto pay higher wages than the solidaristic ones.

83 The importance of solidaristic wage policy for structural change in Sweden shall not beexaggerated. According to a comparative Nordic study, wage equalization was not decisive for the

48

6.2.3 “What´s Good for Volvo is Good Also for Sweden”

The interests of Swedish wage-earners in the Golden Age did not seem to collude withthe expansion of the leading companies. The companies´ support of the Swedish modelwas shown by friendly statements about solidaristic wage and labour market policiesbut also in practice.

Large foreign investments by Swedish companies were not an indication of capitalflight in spite of profit-squeeze tendencies in Sweden. Foreign production decreased inrelation to total foreign sales for the group of leading engineering companies 1966-1973 - see table 11. In fact, for one of the most outward companies, Electrolux, theforeign-units´ share of foreign sales decreased sharply from the early 1960s throughtake-overs of Swedish firms and closures (or sales) of foreign plants.

In general, export and foreign production by large engineering firms were complementsrather than substitutes in the late Golden Age - “closeness” was necessary to satisfyforeign demand. In particular, export from EFTA countries to EU countries wasprevented by trade barriers. In addition, foreign production by the large companiesresulted in export from their Swedish units.

The advantages of being close to foreign markets did not exclude a rivalry whenSwedish companies made decisions whether to produce in Sweden or abroad. Buthere, transportation costs rather than wage costs were decisive. In fact, Swedishcompanies expanded most in countries with high wages and incomes (SOU 1982:27[1982, chapter 6 and 7]). Tendencies of labour scarcity in Sweden, hastened by fullemployment policies, did not induce leading companies to go abroad.

It is true that foreign production increased significantly in relation to exports for earlierhome-market companies such as ASEA and Volvo in the 1960s and early 1970s.Moreover, an increase in the leading engineering companies´ employment abroad inrelation to their total employment - see table 12 - mainly reflected developments inVolvo and ASEA (and Ericsson). But foreign units´ share of foreign sales and totalemployment for ASEA and Volvo was still small in 1973, 38 and 18 per centrespectively for ASEA and 29 and 17 per cent respectively for Volvo.

In addition, head offices and R&D departments in the Swedish transnational companieswere still located in Sweden at the end of the Golden Age. A small increase in foreignunits´ share of R&D expenditures in the late Golden Age for the ompanies togetherreflected one company´s, namely SKFs, establishment of a research centre in theNetherlands in 1972 (SOU 1982:27 [1982, pp. 126-129]) A larger move of strategicactivities from Sweden would have been both an indication and a source of a growingconflict in the Swedish model.

elimination of plants and the production growth in textile and garment, and wood-product sectors inthe 1960s and the first half of the 1970s (Erixon [1984, pp. 27-37]).

49

Employment in Swedish manufacturing decreased both in absolute and relative termsfrom the mid 1960s. The decrease primarily reflected a strong productivity growth inmanufacturing and a demand shift from goods to services, not that manufacturing was“crowded out” by the public sector (for instance, by its financing) or that Sweden hadlost competitiveness due to labour priorities. The loss of market shares for pulpindustries were due to factors which hardly could be blamed on the hegemony oflabour (cf. chapter 7). Besides, the Swedish reduction in the manufactural share ofGDP in the late Golden Age was not dramatic - see table 4.

However, my interpretation of the Swedish model assumes not only an absence ofstrong internal conflicts. It also assumes that capital and labour benefit from eachother. The mutual interests in the Swedish model shall not hide, however, that rawmaterial industries and large engineering companies were more important for theSocial Democratic project than the other way round in the Golden Age.

The prime comparative advantages of Sweden for the leading industrial sector was herresource endowments and the skills and knowledges that had been accumulated in thecountry over a long period on the basis of natural resources and pathbreakinginnovations. Moreover, the attachment of leading managers and owners to Swedenwas probably more based on national affinity than on policy measures.84 It seems alsothat the leading industries were more effective in satisfying labour priorities than thepolicy model itself in the Golden Age.

The small fluctuations in Swedish GDP in the Golden Age (see table 1) reflected astable export growth rather than a successful stabilization policy. In most OECDcountries, including Finland, Iceland and Norway, the variations in export exceeded thevariations in the GDP 1953-1973. This indicates (although roughly), that economicpolicies, together with built-in-stabilisers, evened out business-cycle fluctuations inmany OECD countries. In Sweden, the variations in export and GDP were low and ofequal size. Thus, it was easier for economic-policy makers in Sweden to even outbusiness-cycle fluctuations but their contribution to stabilization was more limitad thanin other countries.85

84 I ignore in this paper the importance of educational and public R & D arrangements for the successof the Swedish model in the Golden Age. Here, much research remains to be done.

85 Swedish stabilization policies in the Golden Age have traditionally been considered as a successstory, at least until the mid 1960s. However, above, I have emphasized the importance of favourableexternal conditions for the stability in Swedish GDP rather than that of economic policies. The swingsin Swedish GDP ought to have been lower than the variations in export as there are strong built-in-stabilisers from import of inputs to Sweden´s differentiated industry and from a rapid public sectorgrowth in the late Golden Age. However, the relatively good balance between inflation andunemployment in Sweden in the early 1970s was primarily due to economic policy. A lower rate ofunemployment than in other OECD countries was obtained by labour market policies, not byKeynesian counter-cyclical policies. The policy mix, however, was more the result of bad timing thana of conscious application of a typical Swedish strategy (cf. The Rehn-Meidner model). Attempts todampen the previous business -cycle peak came (as in the mid 1960s) too late.

50

Charter 7 - Black Clouds At The Horizon - Tensions In The Swedish Model

Above, I have given an idyllic picture of the Swedish model in the Golden Age. Thestate was able to keep up profitability in leading industries and firms which, in theirturn, were able to generate high growth in Sweden. Above all, the Golden Age endswith a profit boom in the leading Swedish industries (1973-1974) primarily a reflectionof high export prices on raw materials and strong investment demand in major OECDcountries. The profit boom also reflected a reluctance of the government to revaluatethe SEK. But there are signs, even in the Golden Age, that the Swedish growth modelhad some shortcomings and costs.

Criticism of the Swedish model formed a “red wave” from the late 1960s. Labourprotests against Taylorism and Fordism, fundamental principles of the model, showedup in strikes, high turn over, shirking and absenteeism. The protests were directlyinduced by rationalizations in leading sectors (including mining) but they reflected alsoan exceptionally strong position of labour in Sweden at the end of the Golden Age.

The “red” revolt against the Swedish model was not only spontaneous. The LO and itstrade unions questioned the prerogative of leading companies to control strategicinvestments and their earnings of “excess profits” generated by the solidaristic wagepolicy. The LO offensive culminated after the profit boom with a radical proposal ofwage-earner funds (1975-1976).86

Wage-earner funds were unanimously resisted by the private business sector inSweden. The resistance shaped the “blue wave” in Sweden, a palpable politicalmovement in almost all Western countries in the 1980s.

Thus, the support of the Swedish model from both capital and labour did graduallyvanish in the 1970s. Labour emphasized the limitations and negative welfare effects ofthe model while capital criticized the attempts to go beyond it and also to push welfarereforms too far. In addition, predecessors of a “green wave” from the mid 1970sdisputed the fundamental priority of growth in the model. The model was accused ofleading to environmental problems (such as dead forests) and regional imbalances.

My account of shortcomings and costs of the Swedish model will concentrate on itsunability to generate sustainable growth, not on its effects on power relations,distribution, work content, regional balances or environment. First, a further expansionof leading industries in Sweden was restricted by some labour priorities. Second,leading companies and industries showed, independently of the priorities of theSwedish model, a weaker growth potential in the late Golden Age. Third, thedominance and strategies of the large companies in engineering had some negativeeffects, although not yet obvious, on Swedish growth, Their foreign orientation and

86 I will not consider the Job Security Act (1974) or the Co-Determination Act (1977) as breaks withthe Swedish model. These laws were rather confirmations than challenges of developments in leadingcompanies.

51

specialization was not unambiguously beneficial for Sweden and the working of theSwedish growth engine prevented the emergence of new growth engines.

7.1 Internal Conflicts in the Model

The tendencies of a decline in pre-tax profitability in the leading industries in theGolden Age were caused by a combination of increasing labour shortage, harderinternational competition and lower demand growth. Economic policies contributed tolabour scarcity since macroeconomic policies were expansive or substituted by labourmarket policies, particularly in the recession of the early 1970s.

Labour shortage led to massive immigration of low skill workers most of them fromFinland. Immigration to Sweden culminated in the economic peaks 1964-1965 and1969-1970. The number of immigrants was, however, not large enough to offset thetendencies of labour shortage and profitability decline in Swedish engineering (with theexception of transport industries less ship building).

According to the French regulation school, labour shortage was not the decisive factorbehind the profitability decline in the Golden Age (Lipietz [1988] and Boyer [1988]).The school claims that productivity growth was hampered by workers´ revolt againstTaylorism and Fordism.87 However, in my view, impoverishing work conditions,related to large-scale production, resulted in growing difficulties to recruit labour forleading industries, not to serious obstacles to productivity growth in Swedishmanufacturing. On the contrary, productivity was stimulated by labour scarcity leadingto rationalizations and the expansion of industries producing labour-saving techniques.

To summarize, “labour strength” in the Golden Age is important in both the Frenchregulation theory and my theory of the profitability decline in Swedish manufacturing.The difference is that strong strong labour impeded productivity growth in the Frenchtheory while it resulted in high wage claims and reluctances to industrial work in mytheory.

The Swedish model was threatened not only by a growing conflict between fullemployment policies and capital accumulation but also, quite independently, bygrowing difficulties of the leading industrial sector to generate high growth in Sweden.

7.2 An Exhaustion of Growth Potentials of the Swedish Model

87 Robert Boyer [1987] adds that productivity potentials were exhausted in the late Golden Age. In theSwedish case, this argument is valid for raw material industries, at least in the medium term. Anexhaustion of productivity potentials explains to a large extent why productivity growth was higher inthe 1960s than in the 1970s in Swedish raw material industries.

52

There were three threats against a further growth of the leading companies andindustries (where markets and technology are concerned) at the end of the GoldenAge.

A saturation of demand for main products.

Low-cost competition inter alia from larger foreign firms andunderdeveloped countries.

Large productivity potentials were, especially in raw material industries,exhausted or, by technological reasons, more difficult to create.

The first two threats are characteristic for “mature” companies particularly if theycome from small countries. They were also threats to Sweden because of thecompanies´ dominant position. At the end of the Golden Age, Swedish manufacturingcould expect a lower demand growth and a fiercer foreign competition relying oneither scale advantages, low wages or low (other) input prices. In addition, it hadbecome easier for underdeveloped countries in the Golden Age to catch-up Westerntechnologies. Knowledge was rapidly spread through trade liberalizations andtransnational companies. Thus, it was more difficult for Swedish industry to counter-balance high nominal costs with productivity advantages.

The raw material industries´ share of Swedish export was sharply reduced in theGolden Age (see table 7). Moreover, the transformation of Swedish export fromunfinished to finished goods was rapid in an international perspective. But, as alreadymentioned, the raw material industries´ shares of employment and production inSwedish manufacturing were surprisingly stable in the Golden Age. Sweden was stillspecialized in wood industries in the mid 1970s although to a somewhat lesser degreethan at the end of World War II.88 Swedish producers were vulnerable since thetransformation from pulp to paper products had been more extensive in competingcountries. Moreover, Swedish pulp producers suffered from high raw material costs incomparison to its North American competitors (Carlsson et al [1979, pp. 95-96]).

Sweden´s specialization in iron and steel products did actually increase in the GoldenAge, thus, in a period when Japan and NIC countries invaded Western markets.External threats to Swedish iron and steel industries were not warded off by up-gradings. Sweden lost world market shares for special-steel products at the end of theGolden Age when earlier quality advantages were captured (LU 80 [1981, pp. 146-178]).

The Swedish change in export composition from semi-finished to finished goods in theGolden Age was not exceptional compared to other raw material (or agricultural) 88 In fact, Sweden increased her market share for wood products (primarily for saw mills) in the1960s and early 1970s in spite of a relatively low world demand (Erixon [1988, pp. 14] and Mjøset etal [1997)]).

53

countries. In fact, there was a clear convergence between OECD countries. Exporttransformation was fastest in countries with the highest share of unfinished goods inthe initial state.89

Sweden did become more specialized in engineering and also in R & D-intensivebranches in the Golden Age. However, the threats of low demand and hardercompetition did also exist in Swedish engineering. Engineering industry was notspecialized in technologically sophisticated products at the end of the Golden Age(Ohlsson [1976] and Carlsson et al [1979, pp. 78-81]). In fact, Ericsson, Saab-Scaniaand ASEA were the only large “high-tech” companies according to conventionalclassifications.90

The threats to established companies and industries from low market growth and“primitive” cost competition did not show up on the macroeconomic level in theGolden Age. However, Swedish growth was kept up (and market shares regained) bysome leading companies and by reconstructions of established industries throughrationalizations (including mergers) and mechanization, not by structural change.Neither case offered promising prospects of sustainable Swedish growth.

There was only one industry which gave a significant contribution to the preservationof Swedish market shares from the mid 1960s - car industry. Besides, the contributionwas primarily a reflection of high market growth for cars, not of higher market sharesfor Swedish car producers (Mjøset et al [1997]). The competitive pressure on Swedishcar industry was raised at the end of the Golden Age through Japanese invasions ofWestern markets and mergers (or take-overs) by European companies. To match scaleadvantages of foreign companies, the manager director of Volvo (P-G Gyllenhammar)and the Wallenberg family tried to merge Saab-Scania and Volvo in 1977. As theyfailed, the Swedish car companies began to penetrate more exclusive market segments.

The potential for productivity increases in Swedish manufacturing was large in the lateGolden Age. It was largely exhausted in raw material industries by rationalizations.Such reconstructions led to a one-time lift in productivity only, not to a stableproductivity growth. Industries in Swedish manufacturing with the highest productivitygrowth 1965-1974 were, in fact, hit by the strongest decline in the latter half of the1970s and the 1980s (Erixon [1990, pp. 73-76]). Significant production slacks inSwedish raw material industries were not built up again until the late 1980s. Beside,mechanization and a maturity of technologies had made it more difficult, even in theGolden Age, to create large productivity potentials by the use of new technologies.

89 Horwitz [1979, p. 43] and Erixon [1984, pp. 25-26].

90 Some steps were taken in the late 1960s to break the tradition of a passive growth policy. Agoverment department for industry and an investment bank was established. State-owned firms wereorganized in a particular industrial group. Finally, a public institution for R&D support to companiesand inventors (STU) was formed. Still, industrial policies played a minor role in the reconstructing ofSwedish manufacturing in the Golden Age. But there are exceptions. R&D supports encouraged,together with public health investment, the expansion of Swedish pharmaceutical and medical-technical industries during the second half of the 1960s, see, for instance, the foundation ofPharmacia Biotech.

54

7.3 The Social Price of Outwardness and Specialization

Hitherto, the existence of outward specialized companies has been seen as acomparative advantage of Sweden. The companies have even been considered as mainactors in a dynamic innovation system. But their enlarged foreign orientation andspecialization in the Golden Age had some negative effects on Swedish growthalthough less apparent in the short run.

The Swedish innovation system might have been impaired by the “core” strategy of thelarge engineering companies in the 1960s. Although successful for the companiesunder discussion, at least in the short run, their avoidance of risky investments in newfields probably reduced the possibilities of Swedish industry to penetrate new productmarkets and also to develop or adapt new technologies.91 The core strategy ledconceivably to a worser growth trajectory for Sweden.

Furthermore, the development of Swedish suppliers in the Golden Age might havebeen restricted by large companies´ export and foreign production. The up-grading ofdomestic suppliers was often neglected by the outward leading companies. Theirforeign investments were partly both a reflection and a cause of a weaker attention todomestic suppliers. The large engineering companies´ unwillingness to take a largerresponsibility for product developments in supplying Swedish firms was serious sincemany suppliers were strongly depending on their orders. The leading companies couldhave speeded up the use and development of new technologies in small and medium-sized Swedish firms and also the concentration of supplying industries.

The foreign orientation of leading companies is one explanation of why Swedishsuppliers, particularly in car industry, did not obtain a high technological level in theGolden Age (cf. Fölster [1991, pp. 216-217]). Furthermore, the large number ofbankruptcies of small firms and the dramatic reduction of employment in Swedishmanufacturing in the early 1990s is partly explained by the vulnerable position ofsuppliers in Swedish engineering (primarily in car industry).

In addition, foreign investments by large Swedish companies could have laid theground for a rivalry between Swedish and foreign units and a shift in the companies´centre of gravity to other countries. It is true that the sharp increase in the companies´foreign production in the Golden Age was corresponded by high export growth.Besides, it did not lead to an unfavourable specialization pattern for Sweden. But themere existence of operating units abroad (whatever the initial reason) opened thepossibility, that production could be moved from one country to another within arelatively short time. Thus, the expansion of Swedish companies abroad might have

91 A “diversification wave”, similar to the American one in the 1950s and the 1960s, started inSwedish engineering industry at the end of the Golden Age. Now, the large Swedish companieswished to avoid a strong dependency on core products. The real starting-shot was Volvo´s buy of thesport company Jofa and Electrolux´ requirement of Facit, a producer of calculators, in 1973. However,the leading companies then went “back to basics” in the 1980s and 1990s. They abandoned productswith weak technological links to the core assortment.

55

founded a rivalry between export and foreign production, particularly where sales tothird countries are concerned.92

The greater possibilities of leading companies to reallocate production betweencountries made managers more inclined in the 1970s to base their production andemployment decisions on temporary cost differentials between countries. As aconsequence, it became more tempting for politicians in Sweden to use measures instabilization policies to improve Swedish competitiveness.

Furthermore, the large Swedish companies´ possession of production units near theirlarge markets in Western Europe and the U.S raised the probability that R & D andother over-head resources were moved from Sweden after the Golden Age. Forinstance, it became more and more important for the large Swedish companies inengineering in the 1970s and the 1980s to design products in close contact with theirforeign customers. The gradual move of R & D activities from Sweden was probably ablow to the Swedish national innovation system as it prevented up-gradings of Swedishunits and reduced the amount of technological spill-over effects produced orassimilated by the leading companies.

7.4 Locking-In Effects

The concept “locking-in effects” is crucial in my discussion of a conflict between theexpansion of the leading industrial sector and other business activities in Sweden. I willuse the concept for either of the following cases:

* The leading companies and industries get resources for a furtherexpansion through their ties to public authorities and major capitalinterests or by the simple reason that they have been profitable orsolid in the past, not necessarily because of promising market prospects.

** The expansion of the leading companies and industries hampersthe development of new growth engines per se.

The second definition of locking-in effects is similar to the one of private crowdingout. The section “new growth engines” has been added to avoid a rather trivialstatement that the expansion of established firms takes resources away from otherfirms when production factors are scarce. My account of “locking-in effects” willcover the case where the development of dynamic firms and industries (with goodprofit prospects in the long run) is prevented by the expansion of established firms and 92 Figures for 1974 and 1978 do not show any increase in third-country sales for an aggregate ofSwedish companies dominated by the leading engineering companies. However, third-country salesincreased to West-Germany because of Volvo´s acquirement of DAF in the Netherlands in 1972 (SOU1982:27 [1982, pp. 240-252]).

56

industries. Thus, I will assume that the expansion of the leading sector has distortingeffects on industrial structure.

My notion of “locking in” is firm-specific (or industry-specific), not technology-specific. In industrial economics, the concept is often used to cover the cases of path-dependent technologies. I will not exclude, that an industrial process is characterizedby locking-in effects in both respects. However, it cannot be taken for granted thatestablished firms use technologies with low capabilities or limited spill-over effects.Analogously, it is not certain that a specialization of leading companies on “mature”products or technologies, as in the Swedish case, excludes the developments ofdynamic industries and firms. I will discuss below whether dynamic industries and firmsin Sweden were crowded out by the leading industrial sector.

7.4.1 The Early Golden Age

At first sight, it seems that renewal occured in the Swedish business sector at least untilthe mid 1960s. (In this chapter, the early Golden Age ends in the mid 1960s and not inthe late 1950s as in chapter 6.) A large number of new firms were established inmanufacturing in the late 1940s. Morover, the number of new establishments increased(also in relation to the total number of firms) from the early 1950s to the mid 1960s.93

But product innovations were neither a driving force behind (as, in fact, in the 1930s),nor a by-product of new establishments in Swedish manufacturing in the early GoldenAge. Entrepreneurship was stimulated by market growth, not the other way round, andmarket growth was determined independently of innovations.94 Formally independentnew firms were not prime exponents of new ideas in the early Golden Age (see theSwedish innovation system). A substantial share of new firms in manufacturing weresubcontractors to leading companies rather than independent actors on world markets.

The growth potential of new firms in the early Golden Age seems to be limited. Newfirms´ contribution to employment in Sweden was almost twice as large in the interwarperiod as in the 1950s and the 1960s.95

High aggregate growth shall not hide that the Swedish model inhibited thedevelopment of new growth engines. The devaluation in 1949, and the absence of a 93 Du Rietz [1980, pp. 21-22] and [1985, pp. 174-175 and 182] and Jagrén [1992, pp. 14-15].However, the new firms´ share of total employment (in the year of entry) was stable from the early1950s to the mid 1960s.

94 The conclusions above are based on a cross-sectional study of new establisments 1954-1968 inplastic and engineering industries (accounting for approximately 45 per cent of employment inSwedish manufacturing). Simultaneity problems might, however, have underestimated the influenceof new firms on growth (Du Rietz [1980, pp. 56, 79-81, 118-123, 160 and 166)]).

95 The shares were clearly below 1 percent in the early Golden Age (Du Rietz [1985, pp. 179-185]).New firms seem to have been more important for employment in manufacturing in the U.S than inSweden (Du Rietz [1980, p. 63] and [1985, pp. 183-184]).

57

revaluation during the Corean War, led to “locking-in effects” in Swedishmanufacturing (cf. Jörberg [1984, pp. 42-44]). The profit boom raised internal fundsfor investments, and therefore the supply of low-cost capital, in established exportindustries, particularly in wood industries. High profits did also raise their share pricesthus crowding out new ventures from the stock market. Moreover, new and smallfirms could not significantly lift up their profit margins after the 1949 devaluation sincethey were seldom foreign orientated or exposed to foreign competition.

In addition, small and new firms were probably crowded out from loan markets in the1950s. They benefited from low rates of interest in the early Golden Age but it seems,that credit rationing disproportionately favoured established firms and sheltered sectorssuch as construction. Locking-in was reinforced by the profit tax system whichfavoured ploughing back of profits. The “double taxation” of profits in Sweden, thustaxes on both companies and share holders, made the financing of investment byretained earnings much cheaper than by new share issues.96 Furthermore, the profit taxsystem led to locking-in effects by favouring established and capital-intensive firms. Itis true that they were disfavoured by revisions of the profit tax system in the late 1940sand the early 1950s. But the large and capital-intensive companies benefited clearlyfrom investment funds from the mid 1950s.

Sweden´s productivity growth in manufacturing and her export performance was notimpressive at all in the 1950s compared to other OECD countries and adjacentdecades.97 In addition, GDP growth was relatively low in Sweden in the 1950s.98 Aweak transformation pressure on established companies, indicated by high averageprofits in manufacturing, is one explanation of the poor Swedish performance. But thesize of the raw material sector was probably a growth-dampening factor taking boththe nature of world demand and the emergence of locking-in effects into account. Rawmaterial producing countries such as Sweden, Norway, Finland and Canada showed aweak overall export performance until the late 1950s.

Locking-in effects in Sweden in the early Golden Age might even have hamperedgrowth in the long run. Few Swedish companies born after World War II have grownto large independent companies.99 Tetra Pak is one of the few examples. The companygot a leading position in inpackaging industry thanks to a Swedish innovation in the

96 Södersten & Lindberg [1983, pp. 25]. The tax discrimination of new share issues still existed inspite of a mitigation of double taxation in the early 1960s through the so called Annell legislation. Itmust be noted, however, that a similar tax favour of retained earnings existed in the U.S. in theGolden Age.

97 See figures on export growth in table 2. Growth in output per employed person in Swedishmanufacturing was lower in the 1950s than in any other decades of the 20th century if the 1900s, the1970s and the 1980s are ignored, Erixon [1991, pp. 245]. Unfortunately, cross-country data onproductivity growth in manufacturing are not available for the 1950s.

98 See table 3. It is true that Swedish GDP growth was relatively low in the late Golden Age as well.But here, the figures for Sweden are weighed down by a rapid expansion of the (public) service sector.(The manufactural share of GDP was stable in the 1950s.)

99 In fact, this is true also for the 1920s and the 1930s.

58

1940s. Gambro, a medical-technical company formed in the 1960s, is another example.None of the companies, however, is today among the 15 largest Swedishmanufacturers where total sales or employment are concerned. Tetra Pak is close butonly after the company´s requirement of Alfa-Laval in 1991.100

Two pharmaceutical companies, Astra and Pharmacia, expanded rapidly in the GoldenAge but they did not reach a position as one of the 15 largest companies inmanufacturing in the Golden Age.101 Today, they are on the list of the 15 largestcompanies.102 However, the size and technological spill-over effects of the Swedishpharmaceutical industry are too limited to justify a notion of growth engine.103

Locking-in effects in the early Golden Age might have impeded the expansion of newindustries in general and the establishment of new firms in particular. New productareas were often penetrated by established companies.104 Their penetration was notonly an indication of locking-in effects. It also reflected a technical and commercialsuperiority by large firms as suggested by the Swedish innovation model. In any case,the growth path might have been more advantageous for Sweden if new industriessuch as the computer and pharmaceutical industries had been invaded by a multifold offirms and owners.

7.4.2 The Late Golden Age

There are clearer indications from the mid 1960s of a weak renewal of the Swedishbusiness sector. Production growth in Swedish manufacturing was relatively low untilthe first oil crisis in spite of a persisting success of the leading companies.105

100 In 1994, Tetra Pak was number 22 and 29 respectively on the list of the largest manufacturingcompanies in Sweden as far as total sales and employment are concerned (Sweden´s LargestCompanies [1995, pp. 1028 and 1032]).

101 Astra, partly owned by the Wallenberg family, based its expansion on a mean for local anaesthetisedeveloped in the late 1940s. Pharmacia, on the other hand, did not expand rapidly until the 1960s.The company was required by Volvo in 1985.

102 In 1973, Astra was number 31 and 28 respectively on the list of the largest Swedish companies inmanufacturing where total sales values and employment are concerned. The 1000 Largest Companiesin Sweden [1974, pp. 108-109]. In 1994, Astra had climbed up to positions 9 and 14 respectively andPharmacia to positions 11 and 11 respectively.

103 It must be noticed, that the “new” companies above were founded before World War II. Pharmaciawas established in 1911 and Astra in 1913. Tetra Pak was formally founded in 1950s but related toÅkerlund & Rausing (as Gambro), formed in the early 1930s.

104 As already noted, jet engines for aircraft were developed by Volvo and ASEA. The latter companyalso became a producer of nuclear power and industrial robots. Furthermore, Saab entered thecomputer and space industry in the mid 1960s (En bok om Saab-Scania [1987, pp. 76-77]). But Saabsold its computer division (Datasaab) to Ericsson in 1981. Ericsson also required Facit in 1983.

105 The large Swedish transnational companies significantly increased their shares of total export ofindustrial goods in the late Golden Age (Blomström & Lipsey [1988, pp. 108-11] and Jagrén [1992,pp. 12-13]).

59

Furthermore, the annual number of new firms decreased from the mid 1960s until thelate 1970s both in manufacturing and in the business sector as a whole (Du Rietz[1985, pp. 174-175] and [1980, pp. 21-22]). It seems that the Swedish decline inentrepreneurship was exceptional, at least in comparison to large OECDcountries.106At the end of the Golden Age, non-agricultural self-employment as aproportion of total civilian employment was lower in Sweden than in all other OECDcountries (OECD [1992b, p. 158]).

Another indication of a weak industrial renewal in Sweden from the mid 1960s wasthat the country experienced losses of market shares due to product inflexibility.Sweden gained market shares for products with low demand growth but lost marketshares for products with high demand growth. Swedish inflexibility emerged not onlyin manufacturing as a whole but also in engineering (Hultén [1988, pp. 188-195 and220] and Mjøset et al [1997]).

The Swedish absence on dynamic world markets might have been a reflection of thelarge companies´ core strategy rather than of the Swedish model as such. Furthermore,the low Swedish production growth (both in manufacturing and the whole economy) inthe late Golden Age was to a large extent caused by labour scarcity which in its turndepended primarily on demographic factors, not on economic policy.107 My hypothesisis, however, that the emergence and development of new industries and independentcompanies was hindered by the Swedish model.

Of course, the Swedish model cannot explain all of the decline in new establishments inthe late Golden Age. The decline in Swedish manufacturing was partly caused by lowerprofitability and demand growth, tendencies which cannot entirely be related to theSwedish model. For instance, some of the lower demand growth for Swedish productsreflected a shift in demand in favour of services. Besides, the priority of public servicesexplains a lot of the low number of new establishments in Sweden. The expansion ofthe public sector in Sweden is partly explained by social attitudes and politicalarrangements not necessarily connected to the Swedish model.

I will insist on, however, that the Swedish model explains a substantial share of thedecline in Swedish entrepreneurship in the late Golden Age. There are two possibletheories. The first theory blames the decline on labour priorities such as rising incomeand wealth taxes, smaller income and wealth differentials, inter alia throughsolidaristic wage policy, and labour legislation.108 The second theory says, on the other

106 Storey [1982, pp. 36-38], Du Rietz [1985, pp. 184-187] and Johnson [1986, pp. 53 and 63-65].

107 The growth of the population from 15 to 64 years old in Sweden was lower 1965-1973 than in anyof the 13 OECD countries Sweden is compared to in the statistical appendix, see OECD, HistoricalStatistics (periodicals).

108 Du Rietz [1980, p. 83-86] and [1985, pp. 187-188], see also Davis & Henrekson [1995]. However,Du Rietz stress in his explanation of the decline in new establishments in Swedish manufacturing inthe late Golden Age the influence of lower demand growth, higher financial entry barriers (caused bystronger scale advantages and mechanization), higher technical entry barriers (a stronger need ofskilled labour) and concentration.

60

hand, that the Swedish accumulation model has prevented the establish and expansionof new firms. In this theory of locking-in effects, the “political” emphasize is laid onthe discriminating role of profit taxes rather than on labour priorities.

The second theory about locking-in effects is probably more valid for Sweden than thefirst one in the late Golden Age. The decline in entrepreneurship happened to early tobe fully explained by wage equalities, public sector growth or labour legislation. Inaddition, the relationship between labour priorities and entrepreneurship isambiguous.109

The profit tax system made it difficult to develop new firms in the late Golden Age.Risk capital remained in large established companies. Besides, the financialconsolidation of these companies made it easier for them to attract equity capital andcredits. It facilitated also their take-over of small companies in engineering in the1960s (cf. Rydén [1971, pp. 198-206]). In addition, low taxes on debts favoured largeestablished firms in the late 1960s since other firms were hit by credit restrictions. Buta theory that new establishments were prevented in the late Golden Age by locking-ineffects cannot be reduced to a theory of biased corporate taxes.

First, the discriminatory role of monetary policies in the late 1960s has already beennoticed. Monetary restraints hit other companies than the leading ones. Second,locking-in effects might have emerged in Sweden through the strong ownership linksbetween the large companies and private banks. Indirectly, the large commercial bankswere major owners of the leading companies. The Wallenberg family is the bestexample of a strong integration between large companies and capital markets inSweden. Besides, other powerful industrial owners, who could have matched theWallenberg family, did not appear in the Golden Age.

The concept “locking-in effects” is too narrow to cover all aspects of a negativerelationship between the expansion of established companies and the formation of newvigorous firms in Sweden in the Golden Age. The lack of new Swedish growth enginesis partly explained by the industrial structure per se.

The large-scale orientation of the leading industrial sector made it more and moredifficult for small companies to develop in Swedish manufacturing. First, the leading(standardized) industries increased their share of total production in manufacturing inthe Golden Age. Second, scale advantages became more important, at least inconsumer-good industries, due to a standardization of consumer tastes, technologicalprogress and a deregulation of trade. Third, fiercer international competition put astronger pressure on Swedish companies to exploit scale advantages. The merger wave

109 It was assumed in an U.S study that entrepreunership is stimulated by high marginal income taxesdue to ease of underreporting income from self-employment compared to wage-salary earnings (Blau,1987, pp. 448). In addition, solidaristic wage policy (and the priority of low-paid wage earners ingeneral) in Sweden might have promoted some entrepreunership in the Golden Age. For instance,high labour costs in wood industries were important conditions for IKEA, an expansive retail-tradecompany in the Golden Age which later became a world success. Finally, welfare and employmentpolicies may reduce the cost of entrepreneurial failure, thus, counter-balancing the stimulating effecton new firms from high rates of unemployment.

61

in the second half of the 1960s led to a sharp rise in firm concentration in Swedishmanufactural branches (Rydén [1971, p. 64] and Rydén & Edberg [1980, p. 196]).

Barriers to entry became higher in Swedish manufacturing in the 1960s due to thegrowing importance of scale advantages (at imperfect capital markets) and firmconcentration. Moreover, entry barriers to dynamic markets were raised by significantR & D expenditures by leading companies. The decline in entrepreneurship in theGolden Age was, to a large extent, a reflection of the large-scale character of Swedishindustry, a pattern already settled before the age of discriminating profit taxes andlabour priorities in politics.Thus, the lack of renewal in manufacturing is basicallyexplained by the logic of the Swedish model.

The Swedish Growth Model - A Summary

62

In this paper, I have regarded the Swedish model as a strong version of a postwaraccumulation model of mutual interests between labour and capital. In the Golden Age,a powerful labour movement accepted private sovereignty in the business sector andpursued policies to support industrial growth. Swedish industry reacted by acceptinglabour priorities and by (strategic) investments in the country.

Similar growth covenants have been analyzed in disparate academic traditions such asthe sociological theory of “historical compromises” and the French regulation school.My contribution to the literature is to distinguish a leading industrial sector in Sweden.Export-orientated raw material industries and large outward companies in engineeringare the Swedish growth engines. I have illuminated the forces behind the leadingsector´s postwar success with a particular emphasis on economic policies.

The success of the large companies in Swedish engineering is explained by their early“outwardness” but also by public orders, economic policies (including devaluations andimport restrictions) and corporate taxes. The companies´ innovative strength arosethrough a dynamic combination of foreign impulses and national systems of innovation.

The leading industrial sector was able to generate strong multiplier effects in theGolden Age. It was also differentiated enough to avoid serious current-account deficitsand fluctuations in economic activity in Sweden. Furthermore, the large companies inengineering acted as effective producers and transmitters of new knowledge to thebenefit of other Swedish companies.

However, there are indications that the growth potential of the Swedish model waslargely exhausted already in the late Golden Age. Leading industries were threatenedby some internal conflicts in the model (basically labour scarcity), market saturation,“primitive” foreign competition and technical restraints. In addition, the worse growthperformance of Sweden from the mid 1970s had its origin in firm strategies andeconomic policies during the Golden Age. Specialization of large companies inengineering might have restricted promising growth trajectories. Furthermore, theirforeign orientation delayed an up-grading of domestic suppliers. The significantincrease in foreign plants made it also easier for the companies to invest abroad at theexpense of investments in Sweden. Finally, the development of new firms andindustries was hampered not only by the profound large-scale character of the Swedishmodel but also by biased corporate taxes and economic policies “locking in” resourcesin established companies and industries.

A question whether the Swedish model is still alive cannot be answered with acertainty of belief. Internationalization and ideological shifts have reduced both theleading companies´ interest in corporatist arrangements and also the general support ofinterventionist policy-instruments, at least on the national level. Finally, the companiesin the Swedish model have been less able to work as effective growth and employmentengines after the Golden Age. Most important, the car producers have been hit byperiodical crises from the mid 1970s.

But it is too early to officially declare the Swedish growth model dead. The leadingcompanies and industries in the Golden Age still dominate Swedish export, industrialproduction and R & D. They have not yet been superseded or complemented by new

63

growth engines. Besides, the support from economic policy to the leading industrialsector increased tremendously after the Golden Age. A free use of investment fundswas permitted from the mid 1970s and the SEK was devaluated at five occasions1976-1982. Moreover, “branches of crisis” such as iron and steel industries gotimmense financial support 1975-1982 and the car industry was given large regionalsubsidies during the peak of the 1980. I argue, in line with my discussion of the GoldenAge, that these supports to leading companies and industries delayed the developmentof new growth engines. It remains, however, to tell the whole story about the Swedishgrowth model after World War II.

Literature

64

Bergström, Villy [1969], Den ekonomiska politiken i Sverige och dess verkningar.Industriens utredningsinstitut, Stockholm.

Bergström, Villy [1981], “Investeringspolitiken under efterkrigstiden”. ArbetsrapportSerie B, nr 1. Nationalekonomiska institutionen, Uppsala Universiet.

Blau, David M [1987], “A Time-Series Analysis of Self-Employment in the UnitedStates”. Journal of Economic Literature, no 3, Vol. 95.

Boyer, Robert [1988], “Technical Change and the Theory of “Regulation””. TechnicalChange and Economic Theory, G Dosi et al (editors), Pinter, London.

Calmfors, Lars [1992], “Lessons from the Macroeconomic Experience of Sweden”.Seminar Paper No. 522, Institute for International Economic Studies, StockholmUniversity, Stockholm.

Carlson, Sune [1979], Swedish Industry Goes Abroad. Studentlitteratur, Lund.

Carlsson, Bo [1972], “The Measurement of Efficiency in Production: An Applicationto Swedish Manufacturing Industries 1968”. The Swedish Journal of Economics, No.4,Vol. 74.

Carlsson, Bo et al [1979], Teknik och industristruktur - 70-talets ekonomiska kris ihistorisk belysning. Industriens utredningsinstitut och Ingenjörsvetenskapsakademin,Almqvist & Wiksell International, Stockholm.

Carlsson, Bo [1988], “Specialisering och teknisk utveckling. Verkstadsindustrin iinternationell jämförelse”. Ledningens visioner - mål & medel, Bratt Publishing,Stockholm.

D´Agostini, Hjördis [1987], “Arbetarförbundens Medlemsutveckling i Sverige 1900-1985”. Meddelande 9/1987, Institutet för social forskning, Stockholm.

Dahmén, Erik [1950], Svensk industriell företagarverksamhet. Band I. Industriensutredningsinstitut.

Dahmén, Erik [1980], “Hur studera industriell utveckling?” Industriell utveckling ISverige - Teori och verklighet under ett sekel. E Dahmén & G Eliasson (redaktörer),Industriens utredningsinstitut, Almqvist & Wiksell International, Stockholm.Stockholm.

Davis, Steven J. & Henrekson, Magnus [1995], “Industrial Policy, Employer Size andEconomic Performance in Sweden”. National Bureau of Economic Research, Inc.,Working Paper No. 5237.

Du Rietz, Gunnar [1980], Företagsetableringarna i Sverige under efterkrigstiden.Industriens utredningsinstitut, Stockholm.

65

Du Rietz, Gunnar [1985], “Nyföretagandets roll för sysselsättningen”.Arbetslöshetsfällan - Strukturproblem och utvecklingsmöjligheter i den svenskaekonomin. Svenska arbetsgivareföreningen, Stockholm.

Dunning, John H [1993]. Multinational Enterprises and the Global Economy.Workingham Addison, Wesley.

Edgren, Gösta, Faxén, Karl-Olof & Odhner, Clas-Erik [1970], Lönebildning ochsamhällsekonomi. Rabén & Sjögren, Stockholm.

En bok om Saab-Scania [1987]. Saab-Scania AB, Streiffert & Co Bokförlag HB,Stockholm.

Englund, Peter [1993], “Bostadsfinansieringen och penningpolitiken”. Frånräntereglering till inflationsnorm - Det finansiella systemet och riksbankens politik1945-1990, SNS Förlag, Stockholm.

Erixon, Lennart [1982], “Why Did Swedish Industry Perform So Poorly in the Crisis”.Skandinaviska Enskilda Banken Quaterly, no 4.

Erixon, Lennart [1984], “Den svenska modellen i medgång - En analys av perioden1945-1973”, Institutet för social forskning, mimeo.

Erixon, Lennart [1985], “What´s Wrong With the Swedish Model?- An Analysis of ItsEffects and Changed Conditions 1974-1985.” Meddelande 12/1995, The Institute forSocial Research, Stockholm.

Erixon, Lennart [1987], Profitability in Swedish Manufacturing- Trends andExplanations. Swedish Institute for Social Research, Stockholm University, Almqvist& Wicksell International, Stockholm.

Erixon, Lennart [1988], “Löner och konkurrenskraft - lönekostnadernas betydelse förSveriges världsmarknadsandelar”. SIND PM 1988:5. Statens industriverk, Stockholm.

Erixon, Lennart [1990], “Produktivitetsnedgången i svensk industri - Fem hypoteser”.Produktivitetsdelegationen, Stockholm.

Erixon, Lennart [1991], “Styrfel - inte systemfel - orsak till Sveriges eftersläpning?”.Ekonomisk Debatt, nr. 1, årg. 19.

Erixon, Lennart [1994], “Investeringar och lönsamhet”. Bilaga 7 till LU 95, Fritzes,Stockholm.

Fagerberg, Jan [1987], “International Competitiveness”. Technology, Growth andTrade (dissertation), Oslo.

Fagerberg, Jan [1994], “Technolgy and International Differences in Growth Rates”.Journal of Economic Literature, No. 3, Vol. 32.

66

Fölster, Stefan [1991], “Hinder för teknikspridning i Sverige”. Forskning,teknikspridning och produktivitet. Expertrapport nr 10 till Produktivitetsdelegationen,Allmänna Förlaget, Stockholm.

Giertz, Eric [1991], Människor i Scania under 100 år - industri, arbetsliv ochsamhälle i förändring. Norstedts ekonomi, Stockholm.

Glete, Jan [1987], Ägande och industriell omvandling. SNS Förlag, Stockholm.

Glete, Jan [1994], Nätverk i näringslivet - ägande och industriell utveckling i detmogna industrisamhället 1920-1990. SNS Förlag, Stockholm.

Gårdlund, Torsten [1947], Svensk industrifinansiering under genombrottsskedet 1830-1913. Svenska bankföreningen, Stockholm.

Gårdlund, Torsten [1955], Industrialismens samhälle. Tidens Förlag, Stockholm.(First edition: 1942.)

Gårdlund, Torsten [1974], “The Finances of Atlas, 1873-1973”. Atla Copco - TheStory of a World-Wide Air Company, Atlas Copco, Stockholm.

Gårdlund, Torsten [1976], Marcus Wallenberg 1864-1943. Norstedt & Söners förlag,Stockholm.

Gårdlund, Torsten [1983], “Förhistoria och uppbyggnad”. Ett världsföretag växerfram - Alfa-Laval 100 år, del I, Ekonomisk historisk forskning (EHF) vidHandelshögskolan i Stockholm, Norstedts Tryckeri, Stockholm.

Hansen, Svend Aage [1970], Early Industrialisation in Denmark. CopenhagenUniversity Institute of Economic History, Copenhagen.

Hansen, Svend Aage [1972], Okonomisk Vaekst i Danmark - bind I: 1720-1914.University Institute of Economic History, Copenhagen.

Harmaja, Leo [1933], Effects of the War on Economic and Social Life in Finland.YaleUniversity Press, New Haven.

Hodne, Fritz [1975], An Economic History of Norway 1815-1970. Tapir, Bergen.

Hoeller, Peter, Louppe, Marie-Odile & Vergriete, Patrice [1996], “Fiscal RelationsWithin the European Union”. Working Papers, no 163, Economics DepartmentOECD, Paris.

Holmlund, Bertil [1990], “Den svenska lönebildningens drivkrafter”. Svensklönebildning - Teori, empiri, politik. Bilaga 24 till Långtidsutredningen 1990,Allmänna Förlaget, Stockholm.

Horwitz, Eva Christina [1979], “Svensk export och internationell handelstillväxt”. IUI40 år - Företagen i marknadsekonomin, Industriens Utredningsinstitut, Stockholm.

67

Hultén, Staffan [1988], Vad bestämmer de svenska exportmarknadsandelarnasutveckling? EFI - Ekonomiska forskningsinstitutet, Handelshögskolan i Stockholm,Stockholm.

Jagrén, Lars [1992], “Det svenska storföretagsberoendet”. Working Paper nr 360,Industriens utredningsinstitut, Stockholm.

Janelid, Ingvar [1974], “Atlas and Compressed Air Technology”. Atlas Copco 1873-1973 - The Story of a World-Wide Compressed Air Company, Atlas Copco,Stockholm.

Johannesson, Jan [1981], On the Composition of Swedish Labour Market Policy.EFA, Arbetsmarknadsdepartementet, Stockholm.

Johnson, Peter [1986], New Firms - An Economic Perspective. Allen & Unwin,London.

Jonung, Lars [1993], “Riksbankens politik 1945-1990”. Från räntereglering tillinflationsnorm - Det finansiella systemet och riksbankens politik 1945-1990, SNSförlag, Stockholm.

Jungenfelt [1966], Löneandelen och den ekonomiska utvecklingen - En empirisk-teoretisk studie. Industriens Utredningsinstitut, Almqvist & Wiksell, Stockholm.

Jörberg, Lennart [1961], Growth and Fluctuations of Swedish Industry 1869-1912.Almqvist & Wiksell, Stockholm.

Jörberg, Lennart [1984], “Den svenska ekonomiska utvecklingen 1861-1983”.Meddelande från Ekonomisk-historiska institutionen, Lunds universitet.

Korpi, Walter [1983], The Democratic Class Struggle. Routledge & Kegan Paul,London.

Korpi, Walter & Palme, Joakim [1993], “Socialpolitik, kris och reformer: Sverige iinternationell belysning”. Bilaga 17 till Nya villkor för ekonomi och politik, bilagedel2, Ekonomikommissionen, SOU 1993:16, Allmänna Förlaget, Stockholm.

Kragh, Börje [1973], “Den svåra stabiliseringspolitiken”. Ekonomisk Debatt, no 3, vol.1.

Laxén-Payro & Odhnoff, Jan [1985], Electrolux - Malmö-Singapore, tur och retur.Arbetslivscentrum, Varia nr 56, Stockholm.

Lindbeck, Assar [1975], Svensk ekonomisk politik - Problem och teorier. Aldus,Stockholm.

Lipietz, Alain [1988], “An Alternative Design for the Twenty First Century”. Optionsin Economic Design. S G Pendse (editor), Greenwood Press, Westport, Connecticut.

68

LU 80 [1981], “Svensk export och import av varor”. Bilaga 7, Ds E 1981:6, Liberförlag, Stockholm.

Lundberg, Erik [1945], Det svenska näringslivets konjunkturkänslighet. Meddelandenfrån Konjunkturinstitutet, Stockholm.

Maddison, Angus [1982], Phases of Capitalist Development. Oxford University Press,Oxford.

Matthiessen, Lars [1971], “Finanspolitiken som stabiliseringspolitiskt instrument”.Svensk finanspolitik i teori och praktik, E Lundberg (editor), Aldus/Bonniers,Stockholm.

Mc Queen, D. H. & Wallmark, J. T. [1983], 100 viktiga innovationer i Sverige 1945-1980. Styrelsen för teknisk utveckling (STU), Stockholm.

Mjøset, Lars et al [1986], Norden Dagen Derpå. De nordiske økonomisk-politiskemodellene og deres problemer på 70- og 80-tallet, Universitetsforlaget ochSamfundslitteratur, Oslo och København.

Mjøset, Lars et al [1997], Comparative Nordic Economic Policy in the 1980s and the1990s. Forthcoming.

Nilsson, Christian [1995], Den interregionala omflyttningen i Sverige. EFAArbetsmarknadsdepartementet, Stockholm.

OECD [1992a], Economic Surveys - Sweden. OECD, Paris.

OECD [1992b], Employment Outlook. July, OECD, Paris.

Ohlson, Lennart [1976], Svensk verkstadsindustris internationella specialisering.Industriens utredningsinstitut, Stockholm.

Olson, Mancur [1990], How Bright Are the Northern Lights? Some Questions AboutSweden. Institute of Economic Research, Lund university, Lund.

Plate, Richard et al [1985], En bok om Volvo. AB Volvo, Göteborg.

Porter, Michael E. [1990], The Competitive Advantage of Nations. The MacMillanPress Ltd.

Ramström, Dick [1974], “Atlas in the World Marketplace”. Atlas Copco 1873-1973 -The Story of a World-Wide Compressed Air Company, Atlas Copco, Stockholm.

Rydeman, Anders [1980], “En historisk studie av de fyra praktikfallen”. Industriellutveckling i Sverige - teori och verklighet under ett sekel. E Dahmén & G Eliasson(redaktörer), Industriens utredningsinstitut, Almqvist & Wiksell International,Stockholm.

69

Rydén, Bengt [1971], Fusioner i svensk industri. Industriens utredningsinstitut,Stockholm.

Rydén, Bengt & Edberg, Jan-Olof [1980],“Large Mergers in Sweden”. TheDetermination and Effect of Mergers - An International Comparison, D C Mueller(editor), Oelgeshlager, Gunn & Hain Publishers Inc., Cambridge Massachusetts.

Schön, Lennart [1990], Electricitetens betydelse för svensk industriell utveckling.Energisystem Utveckling Miljö, Vattenfall, Älvkarleby.

SOU 1982:27 [1982], Svensk industri i utlandet - en analys av drivkrafter ocheffekter. Expertrapport från Direktinvesteringskommittén, Allmänna Förlaget,Stockholm.

SOU 1993:43 [1993], Politik mot arbetslöshet. Betänkande av Expertgruppen förarbetsmarknadspolitiska utvärderingsstudier (EFA), Allmänna Förlaget, Stockholm.

Sejested, Francis [1993], Demokratisk Kapitalisme. Universitetsførlaget, Oslo.

Statens Industriverk [1988], “Sverige - drivhus för internationell konkurrenskraft?”.SIND 1988:1, Stockholm.

Statens Industriverk [1990], “Svensk högteknologisk industri och dess export”. SIND1990:4, Stockholm.

Statistics Sweden [1989], “Forskningsstatistik”. Statistiska meddelanden U14 SM8901, Stockholm.

Statistics Sweden [1990], Offentliga sektorn - Utveckling och nuläge. Sverigesofficiella statistik, Stockholm.

Statistics Sweden [1992], Forskning och utvecklingsarbete i Sverige. Stockholm.

Storey, D.J. [1982], Entrepreunership and the New Firm. Croom Helm, London &Canberra.

Swedenborg, Birgitta [1979], The Multinational Operations of Swedish Firms. AnAnalysis of Determinants and Effects. The Industrial Institute for Economic nd SocialResearch, Stockholm.

Sweden´s Largest Companies [1995], 28th Edition, Ekonomisk Litteratur AB,Stockholm.

Svennilson, Ingvar [1954], Growth and Stagnation in the European Economy.Genève, Förenta Nationerna.

Sveriges Verkstadsförening [1980], Lönsamheten inom verkstadsindustrin 1980.

70

Söderpalm, Sven Anders [1976], Direktörsklubben - Storindustrin i svensk politikunder 1930- och 1940-talen. Tema Teori 12, Zenit/Rabén & Sjögren, Stockholm.

Södersten, Jan [1971], “Företagsbeskattning och resursfördelning”. Svenskfinanspolitik i teori och praktik, Aldus, EFI, Stockholm.

Södersten, Jan & Lindberg, Thomas [1983], Skatt på bolagskapital - Sverige ijämförelse med Storbritannien, USA och Västtyskland. Forskningsrapport nr 20,Industriens utredningsinstitut, Stockholm.

Sölvell, Örjan, Zander, Ivo & Porter, Michael E [1991], Advantage Sweden.Stockholm. In English, 1993, MacMillan Press, London.

The 1000 Largest Companies in Sweden [1974], Veckans Affärer och EkonomiskLitteratur AB, Stockholm.

Warginger, Klas-Göran [1992], Wage Formation in the Scandinavien Model.Department of Economics, Stockholm University.

Wibe, Sören [1976], Monopol och monopolkapitalism? Häften för kritiska studier, no7-8, årg. 9.

Wihlborg, Clas [1993], “Valutapolitiken”. Från räntereglering till inflationsnorm.SNS Förlag, Stockholm.

Visser, Jelle [1989], European Trade Unions in Figures. Netherlands Boston KluwerLaw and Taxation Publishers Co, Deventer.

Wohlin, Lars (1970), “Strukturomvandling inom skogsindustrin”. SkandinaviskaBankens Kvartalsskrift, no. 51.

71

Table 1: Fluctuations in Export and GDP Growth in the Golden Ageand After (Constant Prices), 16 OECD Countries. VariationCoefficients (Per Cent).

1954-1972 1973-1990Export GDP Export GDP

Sweden 42 41 124 72

Denmark 37 57 58 105

Norway 50 45 211 116

Finland(1) 75 51 164 67

Iceland 140 93 142 76

Germany 91 71 129 86

Austria(1) 74 40 72 67

Belgium 72 43 105 93

France(1) 65 22 80 59

Italy(1) 38 32 82 88

Netherlands(1) 40 49 94 82

Great Britain 69 52 98 114

Switzerland 79 77 126 181

USA 127 90 119 90

Canada 74 52 118 78

Japan(1) 51 31 98 47(1) 1955-1973.Sources: National Accounts of OECD Countries, Volume II, OECD (periodical).

72

Table 2: Average Percentage Change In Export Volume in 15 OECD Countries, 1951-1958, 1959-1965 and 1966-1973. Annual Data.

1951-58 1959-65 1966-73

Sweden 5,1 10,9 11,2

Denmark 8,7 9,7 7,9

Norway 5,0 12,3 11,6

Finland 7,8 9,9 10,0

Germany 30,6 12,4 16,2

Austria 20,8 10,7 16,1

Belgium 7,6 15,8 16,4

France 5,1 13,3 15,8

Italy 12,5 27,1 15,1

Netherlands 12,5 12,9 18,9

Great Britain 1,3 4,3 8,7

Switzerland 9,4 10,6 11,9

USA 5,2 7,4 10,5

Canada 4,6 8,9 16,6

Japan 27,5 30,4 21,3

Source: Maddison [1982].

73

Table 3: Average Percentage GDP Growth in 15 OECD Countries, 1950-1958, 1958-1965 and 1965-1973. Annual Data.

1950-1958 1958-1965 1965-1973___________________________________________________

Sweden 3,5 5,9 3,7

Denmark 2,6 6,6 5,0

Norway 3,9 5,1 4,7

Finland 4,7 7,0 6,0

Germany 10,5 7,0 4,9

Austria 7,6 5,3 6,6

Belgium 3,0 5,6 9,1

France 5,2 6,6 6,6

Italy 6,8 6,6 6,5

Netherlands 4,6 6,4 6,1

Great Britain 2,5 4,0 3,4

Switzerland 4,8 6,7 4,5

USA 3,4 5,1 4,3

Canada 5,9 5,9 6,7

Japan 11,1 14,2 15,4

Sources: Maddison [1982].

74

Table: 4 Changes in Value Added (Percentage Points) in Manufacturing asRatio of GDP (Current Prices). 14 OECD Countries, 1960-1965 and 1965-1972.

1960-1965 1965-1972(1) 1960-1972

Sweden -0,8 -2,7 -3,5

Denmark -8,8 -3,0 -11,8

Norway 0 0,6 0,6

Finland -1,3 3,0 1,7

Germany 0,2 -4,5 -4,3

Austria -2,2 -0,1 -2,3

Belgium -0,1 0,1 0

France -0,4 -0,5 -0,9

Italy -1,7 1,7 0

Netherlands -2,6(2) -2,6(3) -5,2

Great Britain -2,1 -2,7 -4,8

USA -0,5 -3,8 -4,3

Canada -0,5 -2,6 -3,1

Japan -2,1 2,7 0,6

Notes: (1) 1973 is an exceptional year and has therefore been excluded here from the Golden Age. (2)1960-1968. (3) 1968-1972.Sources: OECD, Historical Statistics (periodical) and OECD, National Accounts of OECD Countries1962-1979, Volume II (periodical).

75

Table 5: Average Percentage Change in Real Value Added in Manufacturing in 14 OECDCountries, 1960-1965 and 1965-1972. Annual Data.__________________________________________________________________________________

1960-1965 1965-1972(1)______________________________________________

Sweden 6,1 3,8

Denmark 5,8 4,5

Norway 4,6 4,6

Finland 6,0 7,3

Germany 5,7 4,6

Austria 4,3 6,6

Belgium 5,9 6,8

France 7,7 7,7

Italy 7,0 7,2

Netherlands 5,9 6,1

Great Britain 3,1 2,2

USA 6,3 3,1

Canada 7,3 5,3

Japan 12,1 14,4-------------------------------------------------------------------Notes: (1) 1973 is an exceptional year and has therefore been excluded from the Golden Age.Sources: OECD, Historical Statistics (periodical).

76

Table 6: Raw Material Industries´ Share of Gross Value Added (Current Prices) andEmployment in Swedish Manufacturing in the Golden Age.

1951 1956 1961 1966 1970 1973

Value Added

Wood Products 7,9 7,0 7,7 7,7 8,3 8,9

Pulp and Paper 17,1 8,4 6,7 5,9 7,8 7,9

Iron and Steel 5,0 7,5 7,8 6,9 7,9 6,6____________Total 30,0(1) 22,9 21,5 20,5 24,0 23,4

Employment

Wood Products 9,0 8,7 8,0 8,2 8,5 8,7

Pulp and Paper 5,7 6,5 6,8 6,4 6,0 6,2

Iron and Steel 6,5(2) 6,4 6,8 6,9 6,7 6,9____________Total 21,2 21,6 21,6 21,5 21,2 21,8_____________________________________________________________________Notes: (1) An exceptionally high figure during the Corean War. (2) 1952.Sources: Swedish national accounts.

Table 7: The Raw Material Industries´ Share of Swedish Export of Commodities,(Value) in the Golden Age.

1946 1951 1956 1961 1966 1970 1973

Wood Products 11,4 12,3 12,3 8,6 7,2 6,7 7,9

Pulp and Paper(1) 36,6 42,2 28,2 22,6 20,0 17,3 16,1

Iron and Steel(2) 13,8 9,9 14,3 13,6 15,1 15,5 14,1

Total 61,8 64,4 54,8 44,8 42,3 39,5 38,1

77

Notes: (1) Printed matters are excluded. (2) The group includes fabricated metalproducts except machinery and equipment, thus a branch of engineeringaccording to the national accounts.Sources: Trade statistics from Statistics Sweden.

Table 8: Nine Engineering Companies´ Percentage Share of Employment in Swedish Manufacturing, 1966, 1970 and 1973.

1966 1970 1973

AGA n.a 0,5 0,5

Alfa-Laval 1,0 0,5 0,6

ASEA 2,9 3,0 3,3

Atlas Copco 0,5 0,4 0,5

Electrolux 0,6 1,2 1,3(1)

Ericsson 2,1 2,8 2,9

Saab-Scania 2,3(2) 2,6 3,0

SKF 1,4 1,4 1,3

Volvo 2,2(3) 3,4 4,4__________________________________________________________________Total excl. AGA 13,0 15,2 17,4__________________________________________________________________Notes: (1) 1972. (2) The figures for Saab and Saab-Scania have been summed up. (3)

1967.Sources: See table 10.

78

Table 9: Eight Engineering Companies´ Share of Swedish Export ofCommodities (Value) 1966, 1970 and 1973.

1966 1970 1973

Alfa-Laval 1,0 1,3 1,4

ASEA 3,3 3,4 3,2

Atlas Copco 1,4 1,3 1,2

Electrolux 0,7 0,9 1,8

Ericsson 2,3 3,1 3,6

Saab-Scania 1,7(1) 2,7 3,1

SKF 2,5 1,9 1,5

Volvo 5,7 7,3 8,4

Total 18,6 21,9 24,2(1) The figures for Saab and Scania-Vabis have been added up.Sources: see table 10.

79

Table 10: Total Foreign Sales As A Ratio Of Total Sales From 9 SwedishCompanies in Engineering (Value), 1951-1973.

1951 1956 1961 1966 1970 1973

AGA n.a n.a n.a 61(1) 67 67

Alfa-Laval 79 78 71 66(2) 84 85

ASEA n.a n.a n.a 40 44 52

Atlas Copco n.a 79(3) 81 85 89 90

Electrolux n.a 83 77 67(4) 66 71

Ericsson n.a n.a 65 61 73 82

Saab-Scania n.a n.a n.a 26(5) 36 40

SKF n.a n.a n.a 92 92 97

Volvo 24 36 46 52 65 70Total (excl. AGA) 61 67 71

Notes: (1) 1968. (2) 1965. The figure is 85 per cent if Olofström AB, sold in 1969, had been excluded.(Analogously, the figures for 1951, 1956 and 1961 would have been higher if this daughter companyhad been excluded.) (3) 1958. (4) The decrease from 1961 is mainly explained by the requirement ofElectro-Helios (1962) and the exclusion of Electrolux Canada which formally was transfered to anindependent company, Electrolux Corporation (1962). Electrolux sold its minority share (40 per cent)of Electrolux Corporation in 1968 (5) The figures for Saab and Scania-Vabis have been added up.

Sources: Annual Reports, Gårdlund [1973, statistical appendix] and Erixon [1988, pp. 79-81].

80

Table 11: Foreign Affiliates´ Share Of Total Foreign Sales From 9Swedish Companies in Engineering (Value), 1951-1973.

__________________________________________________________________________________1951 1956 1961 1966 1970 1973

AGA n.a n.a n.a n.a n.a. 83

Alfa Laval 64 n.a 78 74 65 65

ASEA n.a n.a n.a 25 26 38

Atlas Copco n.a n.a n.a 63 67 67

Ericsson n.a n.a n.a 59 53 55

Electrolux n.a. 81 82 78 74 67

Saab-Scania n.a n.a n.a 19 26 24

SKF n.a n.a n.a 83 85 86

Volvo 0 0 n.a. 21 26 29______________________________________________________________________Total (excl.AGA) 59 55 54Sources: See table 10.

81

Table 12: The Number Of Employees In Foreign Affiliates In Relation To The TotalNumber Of Employees in 9 Swedish Companies, 1951-1973.

1951 1956 1961 1966 1970 1973

AGA n.a. n.a. 54(1) n.a. 62 61

Alfa -Laval n.a 57(2) 51 46 64(3) 62

ASEA n.a. 12 9 14 17 18

Atlas Copco n.a. 39(4) 46 57 66 70

Electrolux 68 74 70 65 61 64(5)

Ericsson n.a. n.a. 48 50 52 62

Saab-Scania n.a. n.a. n.a. 7(6) 7 9

SKF 68(7) 72 72 78 77 80

Volvo 0 0 2 10(8) 10 17_____________________________________________________________________Total (excl. AGA): 46 45 49

Notes: (1) 1959. (2) 1957. (3) The sharp rise from 1966 is explained by the sell of Olofström AB toVolvo (1969). (4) 1958. (5) 1972. (6) The figures for SAAB and Scania-Vabis have been added up. (7)1952. (8) 1967.Sources: See table 10.

82

Table 13: Average Private Consumption Expenditures (Constant Prices) in Sweden1950-1958, 1958-1963, 1963-1968 and 1968-1973. Annual Data.

1950-58 1958-63 1963-68 1968-73

Electrical Equipment 20,6 6,9 - -

Heating and Cooking - - 5,9 6,2 Appliances, Refrigera- tors, Washing Machines and Similar Major Household Appliances Including Fittings

Sewing Machines and 3,6 3,6 - - Vacuum Cleaners

Wireless and Tele- - - 12,8 17,3 vision Sets and Gramophones

New Purchases of Cars 14,9 15,8 2,6 1,4

Total Private Consumption 2,6 3,9 3,8 1,9Expenditures in theDomestic MarketNotes: The classification of consumption goods was changed resulting in a time series break in 1963.Sources: Statistics Sweden (1975), “Final Consumption Expenditure 1950-1974”. No N 1975:98,appendix 1.

83

Table 14: Percentage Change in Swedish Import in the Golden Age (Value).

1946-48 48-50 50-52 52-54 54-59 59-65 65-70 70-73_____________________________________________________________

Automobiles (1) 19 330 - 7 114 34 145 -18 48

Automobiles, including 81 139 16 83 24 108 3 47components (2)

Electrical Machines, 152 - 6 95 0 118 66 102 32Apparatus andMaterial____________________

Total Import 46 23 47 3 36 81 52 28_____________________________________________________________________________(1) Include trucks and buses.(2) Components of automobiles such as chariots, wheels and chassis.Source: Swedish trade statistics.